id
stringlengths
36
36
url
stringclasses
504 values
title
stringclasses
475 values
section
stringlengths
0
148
content_type
stringclasses
3 values
has_code
bool
2 classes
text
stringlengths
147
22.5k
vector
listlengths
384
384
80602e4f-8d33-559d-bbce-6c4d3c9ca26e
/course/essentials/day-3/sparse-vectors/
Sparse Vectors and Inverted Indexes
Sparse Vector Representations > Comparing Sparse Representations
course
true
Sparse Vectors and Inverted Indexes > Sparse Vector Representations > Comparing Sparse Representations Comparing two sparse representations, you'd be usually interested in how much they agree on the same features/objects (e.g., the same movie rating). The **dot product** distance metric, introduced in the **Day 1** (...
[ 0.09015475958585739, -0.055052608251571655, -0.030177069827914238, -0.018127569928765297, 0.07574576139450073, -0.02016265131533146, 0.0695982500910759, -0.004762172698974609, 0.09825585782527924, -0.021867264062166214, 0.051030658185482025, 0.047179900109767914, 0.05376037210226059, 0.001...
807477e1-17f3-5bb6-a598-5ba76e5f5b34
/documentation/migrate-to-qdrant/
Migration Tool
Installation
documentation
true
Migration Tool > Installation The tool is distributed as a container image and as a standalone binary. ### Container Image The container image runs with Docker, Podman, or any similar container runtime. It supports every source, and the examples in this documentation use it. ```bash docker pull registry.cloud.qdran...
[ 0.05098855867981911, -0.09920741617679596, -0.02130759321153164, -0.011667096056044102, 0.03709868714213371, -0.017941070720553398, 0.06409754604101181, 0.057229891419410706, 0.0580194815993309, -0.011564863845705986, 0.027081571519374847, -0.015616628341376781, 0.06890126317739487, -0.053...
807c25c5-ebac-5895-8ce1-f2c78bf6049b
/documentation/tutorials-basics/reranking-hybrid-search/
Hybrid Search with Reranking
Best Practices in Reranking
documentation
false
Hybrid Search with Reranking > Best Practices in Reranking Reranking with late interaction models can dramatically improve the relevance of search results, especially when combined with hybrid search. Here are some best practices to keep in mind: - **Continuous testing and monitoring**: regularly evaluate your hybrid...
[ 0.027394305914640427, -0.01794964261353016, -0.07671057432889938, -0.054229944944381714, 0.061257652938365936, -0.05799950286746025, 0.05537452921271324, 0.01074519008398056, 0.09050285816192627, -0.014564934186637402, 0.019841469824314117, -0.012669782154262066, 0.06852753460407257, -0.00...
809dceac-14ca-5912-b99c-d06eaee40dbd
/documentation/tutorials-operations/data-streaming-kafka-qdrant/
Kafka Streaming into Qdrant
Installation of Qdrant:
documentation
true
Kafka Streaming into Qdrant > Installation of Qdrant: To install and run Qdrant (self-managed locally), you can use Docker, which simplifies the process. First, ensure you have Docker installed on your system. Then, you can pull the Qdrant image from Docker Hub and run it with the following commands: ```bash docker p...
[ 0.04493320733308792, -0.0687556192278862, -0.04853899031877518, -0.020931461825966835, 0.027442049235105515, -0.020927412435412407, 0.06874815374612808, 0.039636705070734024, 0.06964755803346634, 0.006556578446179628, 0.0117944972589612, -0.01093023456633091, 0.048653215169906616, -0.05987...
80b8bd96-25b4-5f07-9679-7ccb091b0d2e
/documentation/tutorials-basics/search-beginners-local/
Semantic Search 101
6. Ask the Engine a Question
documentation
true
Semantic Search 101 > 6. Ask the Engine a Question Now that the data is stored in Qdrant, you can ask it questions and receive semantically relevant results. ```python hits = client.query_points( collection_name="my_books", query=encoder.encode("alien invasion").tolist(), limit=3, ).points for hit in hit...
[ 0.047121334820985794, -0.03550919517874718, -0.03471982479095459, -0.048251744359731674, 0.052955787628889084, -0.02497890777885914, 0.03177263215184212, 0.027960877865552902, 0.08872950822114944, -0.047013670206069946, 0.03510839864611626, 0.038133274763822556, 0.07401631027460098, -0.012...
80daf5dc-bab2-5a0c-9418-a9f9a66dcd3e
/documentation/scaling/consistency-guarantees/
Consistency Guarantees
Write Consistency Factor
documentation
true
Consistency Guarantees > Write Consistency Factor The `write_consistency_factor` represents the number of replicas that must acknowledge a write operation before responding to the client. It is set to 1 by default. It can be configured at the collection's creation or when updating the collection parameters. This valu...
[ 0.0004911924479529262, -0.035023435950279236, -0.06817708164453506, -0.011057374998927116, 0.04297734051942825, -0.026160771027207375, 0.031209858134388924, 0.039595168083906174, 0.01130681298673153, -0.03187211975455284, 0.024535570293664932, 0.01562896929681301, 0.032368253916502, -0.033...
80dd468e-d095-5557-999d-32578fb1ed25
/documentation/data-synchronization/with-postgres/
Keeping Postgres and Qdrant in Sync
Tier 1: Application-Level Dual-Write > Where It Falls Apart
documentation
false
Keeping Postgres and Qdrant in Sync > Tier 1: Application-Level Dual-Write > Where It Falls Apart The write path is coupled to Qdrant availability. If Qdrant has a hiccup — even a brief one — you're generating drift that has to be cleaned up later. There's no guarantee that every write will reach Qdrant; you're relyin...
[ 0.008924133144319057, -0.058406148105859756, -0.04140717163681984, -0.028914334252476692, 0.05034894496202469, -0.06349445134401321, 0.0757451131939888, 0.02668147347867489, 0.08427170664072037, -0.0006375557859428227, 0.03600959852337837, 0.020388159900903702, 0.04529890418052673, 0.00025...
811acddc-c78b-5ed4-8c42-4ba32b123693
/documentation/search-patterns/choose-embedding-model/
How to Choose an Embedding Model: Evaluation & Tradeoffs
Balancing All Aspects
documentation
false
How to Choose an Embedding Model: Evaluation & Tradeoffs > Balancing All Aspects After all these considerations, you should have a table that summarizes each of the models you evaluated under all the different conditions. Now things are getting hard and answers are not obvious anymore. Here's an example of how such ...
[ 0.015493184328079224, -0.041259847581386566, -0.03506511077284813, -0.02767932042479515, 0.05196598172187805, -0.029919587075710297, 0.06675446778535843, 0.03383849188685417, 0.06864138692617416, -0.061649393290281296, 0.028653966262936592, 0.003154605161398649, 0.05958223342895508, -0.050...
8131a526-acab-55ad-966c-9ed78dd335bb
/documentation/tutorials-basics/reranking-hybrid-search/
Hybrid Search with Reranking
Implementation > Rerank
documentation
true
Hybrid Search with Reranking > Implementation > Rerank The hybrid search results can be reranked using late interaction embeddings for maximum precision. Instead of fusing with RRF, use the ColBERT multi-vector as the final ranking signal: ```python prefetch = [ models.Prefetch( query=models.Document(text...
[ 0.05007972940802574, -0.032148201018571854, -0.03874531760811806, -0.04335246607661247, 0.08965587615966797, -0.034317947924137115, 0.07929959893226624, 0.043921612203121185, 0.0805918350815773, 0.008466103114187717, 0.04262709990143776, 0.003731133881956339, 0.05514679104089737, -0.024444...
8138eb4c-1b16-5a17-ad8a-e74353446b1a
/articles/neural-search-tutorial/
Neural Search 101: A Complete Guide and Step-by-Step Tutorial
Which model could be used?
article
false
Neural Search 101: A Complete Guide and Step-by-Step Tutorial > Which model could be used? It is ideal to use a model specially trained to determine the closeness of meanings. For example, models trained on Semantic Textual Similarity (STS) datasets. Current state-of-the-art models can be found on this [leaderboard](h...
[ 0.04381806403398514, -0.042990345507860184, -0.037151772528886795, 0.012947553768754005, 0.043183181434869766, -0.0544729009270668, 0.045702267438173294, 0.023260770365595818, 0.042849645018577576, -0.010272794403135777, 0.05591452866792679, 0.020313668996095657, 0.07564559578895569, -0.05...
8157b5ee-e865-5dd5-a87f-0daa5f52cdc6
/course/essentials/day-3/hybrid-search/
Hybrid Search and the Universal Query API
Reciprocal Rank Fusion (RRF) > RRF Example: Step by Step
course
false
Hybrid Search and the Universal Query API > Reciprocal Rank Fusion (RRF) > RRF Example: Step by Step Let's assume we have these results coming from both methods: **Dense Search Results:** 1. D1 (score: 0.95) 2. D2 (score: 0.89) 3. D3 (score: 0.85) 4. D4 (score: 0.82) **Sparse Search Results:** 1. D5 (score: 15.2) 2....
[ 0.055741459131240845, -0.003917114343494177, -0.08449376374483109, -0.03373533859848976, 0.09596598148345947, -0.02414986491203308, 0.04984981566667557, 0.03837300464510918, 0.07625105232000351, -0.03835582360625267, 0.038651347160339355, 0.010132106952369213, 0.048646990209817886, -0.0024...
816d38ba-341c-5285-b4ab-1d357f43d163
/articles/search-feedback-loop/
Relevance Feedback in Information Retrieval
Dismantling the Relevance Feedback
article
false
Relevance Feedback in Information Retrieval > Dismantling the Relevance Feedback Both industry and academia tend to reinvent the wheel here and there. So, we first took some time to study and categorize different methods — just in case there was something we could plug directly into Qdrant. The resulting taxonomy isn'...
[ 0.04055081680417061, -0.018937205895781517, -0.043268755078315735, -0.057970091700553894, 0.04872800409793854, -0.036217667162418365, 0.044505488127470016, 0.036287080496549606, 0.07541538029909134, -0.026965593919157982, 0.039937928318977356, 0.03234107047319412, 0.07346747815608978, -0.0...
817ebf72-ca5c-5ca4-9705-d2a1d04dc2ca
/documentation/faq/qdrant-fundamentals/
Qdrant Fundamentals
Search > My search results contain vectors with null values. Why?
documentation
false
Qdrant Fundamentals > Search > My search results contain vectors with null values. Why? By default, Qdrant tries to minimize network traffic and doesn't return vectors in search results. But you can force Qdrant to do so by setting the `with_vector` parameter of the Search/Scroll to `true`. If you're still seeing `"v...
[ 0.02428232692182064, -0.06708848476409912, -0.0525113008916378, -0.015400202944874763, 0.04269229993224144, -0.016434496268630028, 0.06996957957744598, 0.04847647622227669, 0.07995469868183136, -0.0020872927270829678, 0.038281530141830444, 0.027154913172125816, 0.0637921541929245, -0.01279...
819227cb-86d9-529f-808f-9cd40add598f
/documentation/tutorials-search-engineering/index-dynamic-payloads/
Indexing Payloads of Random Shape
Tie Key to Value With a Nested Filter
documentation
true
Indexing Payloads of Random Shape > Tie Key to Value With a Nested Filter Filtering `attrs[].key` as `"refresh_rate"` and `attrs[].value` as `"144"` with two independent conditions has a trap: Qdrant matches a point where *some* element has that key and *some* element has that value, not necessarily the same element. ...
[ 0.07255934923887253, -0.029969224706292152, -0.06974304467439651, -0.03319158777594566, 0.038180626928806305, -0.012434355914592743, 0.030753320083022118, 0.026700014248490334, 0.08859507739543915, -0.02664010412991047, 0.01889626309275627, 0.02877410501241684, 0.06713717430830002, 0.00967...
819a9660-462f-59c7-a47d-6fe3e83ea9f5
/articles/product-quantization/
Product Quantization for Vector Search
Choosing a quantization method
article
false
Product Quantization for Vector Search > Choosing a quantization method Qdrant now offers four quantization methods. The [quantization documentation](/documentation/manage-data/quantization/#how-to-choose-the-right-quantization-method) maps a target compression to a method: | Compression | Method | |---|---| | 4x | [...
[ 0.03259490057826042, -0.046443816274404526, -0.06606908142566681, -0.04019348323345184, 0.025575120002031326, -0.009871216490864754, 0.06305933743715286, 0.03848440572619438, 0.12795493006706238, -0.01883821375668049, 0.03153305873274803, 0.033998314291238785, 0.02639455907046795, -0.03293...
81ac694b-7f98-512a-a431-5ed2e1c99804
/documentation/private-cloud/api-reference/
API Reference
qdrant.io/v1 > Resource Types > KubernetesStatefulSet
documentation
false
API Reference > qdrant.io/v1 > Resource Types > KubernetesStatefulSet _Appears in:_ - [QdrantClusterSpec](#qdrantclusterspec) | Field | Description | Default | Validation | | --- | --- | --- | --- | | `annotations` _object (keys:string, values:string)_ | Annotations specifies the annotations for the StatefulSet. | |...
[ 0.06390506029129028, -0.04658444970846176, -0.015399567782878876, -0.06919350475072861, 0.02971821278333664, -0.06773465126752853, 0.0475262850522995, 0.03746079280972481, 0.03110077977180481, -0.029118970036506653, 0.02893887646496296, 0.0016776156844571233, 0.03963663801550865, -0.039343...
81bac44c-0ac9-5a6d-86a7-c995778d4240
/documentation/manage-data/payload/
Payload
Payload types > Keyword
documentation
true
Payload > Payload types > Keyword `keyword` - string value. Example of single and multiple `keyword` values: ```json { "name": "Alice", "friends": [ "bob", "eva", "jack" ] } ```
[ 0.0077168140560388565, -0.017254041507840157, -0.03342432528734207, -0.07720573246479034, 0.05713233724236488, -0.03719771280884743, 0.037531159818172455, 0.03268927335739136, 0.07909698784351349, -0.020838122814893723, 0.003003349294885993, -0.024147681891918182, 0.06718181073665619, -0.0...
81e27ec6-22b6-5ddb-a9c8-9f3bd3f6f93a
/documentation/manage-data/collections/
Collections
Collection Info > Grey Collection Status
documentation
true
Collections > Collection Info > Grey Collection Status _Available as of v1.9.0_ A collection may have the grey ⚫ status or show "optimizations pending, awaiting update operation" as optimization status. This state is normally caused by restarting a Qdrant instance while optimizations were ongoing. It means the colle...
[ 0.015222949907183647, -0.038436710834503174, -0.0005270008696243167, -0.0745895579457283, 0.07446055114269257, -0.07283632457256317, 0.04782313480973244, 0.021461406722664833, 0.08851610869169235, -0.07361558079719543, 0.02122536487877369, 0.007737681269645691, 0.05670986324548721, -0.0669...
822d72cf-0ced-5b09-a8f6-d36b9d4f4c74
/documentation/data-management/airbyte/
Airbyte
Setting up Qdrant as a destination
documentation
false
Airbyte > Setting up Qdrant as a destination Once you have a running instance of Airbyte, you can set up Qdrant as a destination directly in the UI. Airbyte's Qdrant destination is connected with a single collection in Qdrant. ![Airbyte Qdrant destination](/documentation/frameworks/airbyte/qdrant-destination.png) ##...
[ 0.04967917501926422, -0.03866741806268692, -0.007602838333696127, -0.04600149393081665, 0.07029393315315247, -0.041182417422533035, 0.03806213289499283, 0.06462416797876358, 0.034724555909633636, -0.047268256545066833, 0.04458269849419594, 0.007781077641993761, 0.014203628525137901, -0.026...
822d97b3-529c-5010-b63a-c96d09ed96f3
/documentation/data-synchronization/with-postgres/
Keeping Postgres and Qdrant in Sync
Tier 1: Application-Level Dual-Write > What This Approach Gets Right
documentation
false
Keeping Postgres and Qdrant in Sync > Tier 1: Application-Level Dual-Write > What This Approach Gets Right The strongest argument for dual-write isn't correctness — it's cognitive simplicity. A new engineer can read this code and immediately understand the entire sync story. There are no background workers, no queues,...
[ -0.01167963445186615, -0.03171950951218605, -0.057823870331048965, -0.015282358042895794, 0.019665373489260674, -0.033383939415216446, 0.06714072078466415, 0.03240009397268295, 0.04787272587418556, 0.026262661442160606, 0.027333542704582214, 0.04777832329273224, 0.060688719153404236, -0.01...
823bb499-efb2-5cd6-b237-9cb98b25c334
/documentation/tutorials-operations/incremental-embedding-updates/
Incremental Embedding Updates
Conclusion
documentation
false
Incremental Embedding Updates > Conclusion A deterministic ID, a content fingerprint, and five sync cases keep embeddings in sync with changing raw data, re-embedding only what actually changed. Adapt this pipeline to your own documents. Ways to make it better: - Pipelines that risk concurrent iterative updates: see...
[ 0.030011001974344254, -0.021775150671601295, -0.05422737076878548, -0.07325150817632675, 0.04811272397637367, -0.017316000536084175, 0.06630551069974899, 0.024227922782301903, 0.05140462890267372, -0.018262628465890884, -0.010888070799410343, 0.029224371537566185, 0.03661549463868141, -0.0...
82586408-df66-5031-adca-3fd09b1f1284
/documentation/tutorials-search-engineering/turbo4-multivector-search/
Compressed Multivector Search
Upload Data
documentation
true
Compressed Multivector Search > Upload Data With the collection created, upload the data and let Cloud Inference embed it server-side, so you never load an embedding model locally. ```python import uuid from typing import Any def get_image(img_dict: dict[str, Any]) -> str: try: return img_dict["large"] ...
[ 0.047117382287979126, -0.021183324977755547, -0.04516021907329559, -0.08475445955991745, 0.08091368526220322, 0.003884243546053767, 0.07861886918544769, 0.029510313645005226, 0.052653733640909195, -0.0022252954076975584, 0.012565342709422112, 0.037602126598358154, 0.059864532202482224, -0....
8269fa94-ba4b-562c-8bdb-fd8e8eae4ad7
/documentation/search/explore/
Explore
documentation
false
Explore After mastering the concepts in [search](/documentation/search/search/), you can start exploring your data in other ways. Qdrant provides a stack of APIs that allow you to find similar vectors in a different fashion, as well as to find the most dissimilar ones. These are useful tools for recommendation systems...
[ 0.04844362288713455, -0.07589524984359741, -0.03790915757417679, -0.031777285039424896, 0.06540258228778839, 0.000052149840485071763, 0.08045817166566849, 0.03976715728640556, 0.0626094713807106, -0.024332651868462563, 0.020358536392450333, 0.052173495292663574, 0.08424929529428482, -0.012...
8297c20f-e999-5dc6-bb7d-47231d397951
/documentation/send-data/databricks/
Databricks Ingestion
Vectorize the Data
documentation
true
Databricks Ingestion > Vectorize the Data Generate both dense and sparse vectors using [FastEmbed](https://qdrant.github.io/fastembed/) in a partition function. Each partition initializes its own models. Workers need network access to download the model files on first use. Both models embed `sentence1`, so the two ve...
[ 0.07556281238794327, -0.026654502376914024, -0.013148334808647633, -0.05470496043562889, 0.07273886352777481, -0.037416938692331314, 0.05569218471646309, 0.06508036702871323, 0.026565561071038246, -0.030818434432148933, 0.025766761973500252, 0.06959442049264908, 0.027659909799695015, -0.03...
829dce5e-653c-526f-9822-51a43e228fed
/documentation/private-cloud/logging-monitoring/
Logging & Monitoring
Logging
documentation
true
Logging & Monitoring > Logging You can access the logs with kubectl or the Kubernetes log management tool of your choice. For example: ```bash kubectl -n qdrant-private-cloud logs -l app=qdrant,cluster-id=a7d8d973-0cc5-42de-8d7b-c29d14d24840 ``` **Configuring log levels:** You can configure log levels for the databa...
[ 0.00011883366823894903, -0.06731836497783661, -0.026358874514698982, -0.02427143231034279, 0.044066622853279114, -0.06662581115961075, 0.025930006057024002, 0.06721197813749313, 0.04559294506907463, -0.0339561328291893, 0.014905793592333794, 0.01789719983935356, 0.05420157313346863, -0.023...
82a11175-cb8f-509b-b88e-098538cfb249
/documentation/overview/
Overview
Deployments
documentation
false
Overview > Deployments | Feature | Benefits | OSS | Managed | Hybrid | Private | | :---- | :---- | :---: | :---: | :---: | :---: | | Deployment | Choose how and where to deploy your Qdrant vector database based on your infrastructure needs. | ✅ | ✅ | ✅ | ✅ | | High Availability | Automatic failover and replication to ...
[ 0.01914636418223381, -0.05117827281355858, -0.06279629468917847, -0.019040487706661224, 0.08710387349128723, -0.031659044325351715, 0.061903081834316254, 0.049868881702423096, 0.04457985982298851, -0.036389950662851334, 0.025927482172846794, 0.004869722295552492, 0.017017193138599396, -0.0...
82a24b0a-f2d0-518c-8b0e-ce204e9a4f86
/articles/metric-learning-tips/
Metric Learning Tips & Tricks
Training the model
article
false
Metric Learning Tips & Tricks > Training the model There are several ways to train a metric learning model. Among the most popular is the use of Triplet or Contrastive loss functions, but I will not go deep into them in this article. However, I will tell you about one interesting trick that helped us work with unifie...
[ 0.052352894097566605, -0.02757776714861393, -0.056617461144924164, -0.008095784112811089, 0.07764940708875656, -0.004345841705799103, 0.0501471646130085, 0.025667285546660423, 0.06617172807455063, -0.015660418197512627, 0.03290592133998871, 0.0018603447824716568, 0.07005332410335541, -0.04...
82bc4c8a-7602-564f-baca-74a328f66e22
/course/multi-vector-search/module-1/problems-multi-vector/
Problems of Multi-Vector Search
The Resource Challenge: Memory and Computational Overhead
course
false
Problems of Multi-Vector Search > The Resource Challenge: Memory and Computational Overhead Beyond indexing challenges, multi-vector search demands significantly more resources than traditional single-vector approaches. You've seen the precision benefits in the previous lesson on use cases - now let's understand the c...
[ 0.04208087921142578, -0.04088757559657097, -0.048653583973646164, -0.054214417934417725, 0.061159368604421616, -0.042952194809913635, 0.05070749670267105, 0.010760136879980564, 0.10954652726650238, -0.04202060028910637, 0.015136320143938065, -0.003682635957375169, 0.036435566842556, -0.045...
82c35854-bd21-51ac-ae87-1e82b89048ed
/documentation/migrate-to-qdrant/from-elasticsearch/
From Elasticsearch
Concept Mapping
documentation
false
From Elasticsearch > Concept Mapping | Elasticsearch | Qdrant | Notes | | :--- | :--- | :--- | | Index | Collection | One-to-one mapping | | Document | Point | Each document becomes a point | | `dense_vector` field | Vector | Mapped automatically | | Document fields | Payload | Non-vector fields become payload | | `co...
[ 0.07738354802131653, -0.05825265496969223, -0.07422880083322525, -0.05326973646879196, 0.06155936419963837, -0.021260133013129234, 0.07787026464939117, 0.050669942051172256, 0.0679597556591034, -0.04354352504014969, 0.029443953186273575, 0.010784861631691456, 0.02985880896449089, -0.022283...
83156f93-3672-5414-b36e-9b7437e98068
/documentation/cloud/create-cluster/
Create a Cluster
Create a Cluster
documentation
false
Create a Cluster > Create a Cluster ![Create Cluster Page](/documentation/cloud/create-cluster.png) This page shows you how to use the Qdrant Cloud Console to create a custom Qdrant Cloud cluster. > **Prerequisite:** Please make sure you have provided billing information before creating a custom cluster. 1. Start i...
[ 0.011548634618520737, -0.009916639886796474, -0.04784860089421272, -0.04551885277032852, 0.04938049241900444, -0.041669752448797226, 0.037427425384521484, 0.043436449021101, 0.05567562207579613, -0.04108592867851257, 0.04456683620810509, -0.003419734537601471, 0.05968353524804115, -0.04069...
834a135c-5795-5df4-8a1e-ba490640d230
/documentation/tutorials-search-engineering/collaborative-filtering/
Collaborative Filtering
Implementation > Setup
documentation
true
Collaborative Filtering > Implementation > Setup Before you start, you need: - A Qdrant instance: a free [Qdrant Cloud](https://cloud.qdrant.io/) cluster or a local one. Set `QDRANT_HOST` and `QDRANT_API_KEY`. - Python 3.10 or later, with these packages: ```bash pip install qdrant-client pandas ``` You have to firs...
[ 0.036623481661081314, -0.017595205456018448, -0.01774642802774906, -0.02582394704222679, 0.07656488567590714, -0.04329785704612732, 0.04171358793973923, 0.041582051664590836, 0.06038356199860573, -0.028164785355329514, 0.014867828227579594, 0.011126479133963585, 0.07001756131649017, -0.034...
83582bb1-c0c6-51bb-a4b5-9f80417a1446
/documentation/scaling/vertical-scaling/
Vertical Scaling
When Vertical Scaling Is No Longer Enough
documentation
false
Vertical Scaling > When Vertical Scaling Is No Longer Enough These signals mean it's time to scale horizontally instead of resizing further: - Your data volume exceeds what a single node can hold, even with quantization. - CPU on your largest available node size is already maxed out with unacceptable query latency. -...
[ 0.015297546051442623, -0.027517125010490417, -0.03102753683924675, -0.03699477016925812, 0.08526480942964554, -0.021787740290164948, 0.07224531471729279, 0.013178858906030655, 0.09518345445394516, -0.03275468945503235, 0.02054430916905403, -0.015949349850416183, 0.04627443850040436, -0.004...
8393afbb-1b6a-5fa1-b7db-df790e07e16a
/documentation/tutorials-operations/gpu-accelerated-hnsw-indexing/
GPU-Accelerated HNSW Indexing
GPU vs. CPU Comparison
documentation
false
GPU-Accelerated HNSW Indexing > GPU vs. CPU Comparison Both clusters ran Qdrant v1.19, had identical specs (16 GB RAM, 4 vCPU, 64 GB disk), and indexed the same 100,000 vectors with the same `m` and `ef_construct`, so the only variable between the two runs was the presence of a GPU. These figures are specific to this ...
[ 0.022421374917030334, -0.047061070799827576, -0.07864339649677277, -0.012885870411992073, 0.06407193094491959, -0.027635471895337105, 0.0637316033244133, 0.019997499883174896, 0.09298282116651535, -0.028772136196494102, 0.053881339728832245, -0.0016581510426476598, 0.06292112171649933, 0.0...
83a4d8b3-8950-5c64-b4d2-ac1de8bcc414
/documentation/send-data/databricks/
Databricks Ingestion
Upload the Data to Qdrant
documentation
true
Databricks Ingestion > Upload the Data to Qdrant Choose a new collection name and replace `{collection_name}` in the HTTP requests with that name. Use the same name in the upload options. - **Create a Qdrant collection:** - [Create a new collection](/documentation/manage-data/collections/#create-a-collection) using...
[ 0.06900909543037415, -0.016169557347893715, -0.04587666690349579, -0.015984060242772102, 0.07266072183847427, -0.04765934497117996, 0.05731184780597687, 0.09262673556804657, 0.05438026040792465, -0.05291585996747017, 0.021278206259012222, 0.0397181399166584, 0.030203046277165413, -0.020556...
83aa0e3d-ef85-5f01-a951-a8798b98d7c0
/documentation/support/
Support
Qdrant Cloud Support > Ticket Information
documentation
false
Support > Qdrant Cloud Support > Ticket Information Please provide as much detail as possible when submitting your request. This helps us help you faster. This includes but is not limited to: - The ID of your Qdrant Cloud cluster (auto-filled if possible) - Which collection(s) are affected - Code examples showing ho...
[ 0.02581390179693699, -0.03741224482655525, -0.032289642840623856, -0.04923548921942711, 0.05895467847585678, -0.038678932934999466, 0.04367424175143242, 0.06206265836954117, 0.06022052839398384, -0.0578753761947155, 0.017980776727199554, 0.012732908129692078, 0.015334309078752995, -0.04193...
83f7f7e4-6de8-534a-888e-fdacf80289b5
/documentation/manage-data/indexing/
Indexing
Full-Text Index > Stopwords
documentation
true
Indexing > Full-Text Index > Stopwords Stopwords are common words (such as "the", "is", "at", "which", and "on") that are often filtered out during text processing because they carry little meaningful information for search and retrieval tasks. In Qdrant, you can specify a list of stopwords to be ignored during full-...
[ 0.04112889990210533, -0.054770082235336304, -0.01950760930776596, -0.03529435396194458, 0.022320162504911423, -0.025798212736845016, 0.047878697514534, 0.02732161059975624, 0.08484139293432236, -0.02592656761407852, 0.049021679908037186, 0.03474355489015579, 0.0499674454331398, -0.00148733...
83fdb923-01b4-52a5-9897-59721b6c0f2e
/articles/what-is-a-vector-database/
What is a Vector Database?
The Architecture of a Vector Database > RAM-Based and Memmap Storage
article
true
What is a Vector Database? > The Architecture of a Vector Database > RAM-Based and Memmap Storage By default, Qdrant stores vectors in RAM, delivering incredibly fast access for datasets that fit comfortably in memory. But when your dataset exceeds RAM capacity, Qdrant offers Memmap as an alternative. Memmap allows y...
[ 0.04857993125915527, -0.053771741688251495, -0.03317681699991226, -0.057960882782936096, 0.06134219840168953, -0.03633485734462738, 0.07221561670303345, 0.035085853189229965, 0.06311804056167603, -0.022678181529045105, 0.0077218892984092236, -0.00771971233189106, 0.059944361448287964, -0.0...
841824cd-c2e5-58d8-8674-a6d334876723
/articles/modern-sparse-neural-retrieval/
Modern Sparse Neural Retrieval: From DeepCT to SPLADE++
Key Takeaways: When to Choose Sparse Neural Models for Retrieval
article
false
Modern Sparse Neural Retrieval: From DeepCT to SPLADE++ > Key Takeaways: When to Choose Sparse Neural Models for Retrieval Sparse Neural Retrieval makes sense: - In areas where keyword matching is crucial but BM25 is insufficient for initial retrieval, semantic matching (e.g., synonyms, homonyms) adds significant val...
[ 0.014975731261074543, -0.05193188041448593, -0.017154322937130928, -0.040067415684461594, 0.04554729163646698, -0.05196797102689743, 0.06412757933139801, 0.0523003414273262, 0.07357341051101685, -0.01698271371424198, 0.0071026477962732315, 0.05657189339399338, 0.03889720141887665, -0.01920...
842c01ff-4068-5c0e-8a8f-4e45ecb29994
/articles/dedicated-vector-search/
Why Vector Search Needs a Dedicated Database
Summing up
article
false
Why Vector Search Needs a Dedicated Database > Summing up When it comes to vector search, there's a clear distinction between using a dedicated vector search solution and extending a database to support vector operations. **For small-scale applications or prototypes handling up to a million data points, a non-optimiz...
[ -0.008033749647438526, -0.04323926568031311, -0.05783214420080185, -0.034749630838632584, 0.03137744218111038, -0.040461763739585876, 0.057576846331357956, 0.008937301114201546, 0.05671451985836029, -0.032771266996860504, 0.013638083823025227, 0.027432287111878395, 0.04263848066329956, -0....
8435a611-013c-5312-bdbf-678468df6891
/course/essentials/day-7/superlinked/
Integrating with Superlinked
course
false
Integrating with Superlinked Day 7 Advanced feature engineering for vector search applications. ## What You'll Learn - Advanced feature engineering techniques - Vector space optimization - Multi-modal data handling - Performance enhancement strategies ## Mixture of Encoders Architecture in Superlinked The Mixture...
[ 0.036821749061346054, -0.027013808488845825, -0.0845126211643219, -0.04173364117741585, 0.05486183986067772, -0.0326668880879879, 0.041157931089401245, 0.04582327604293823, 0.08109501749277115, -0.011127978563308716, -0.0026186397299170494, 0.006878114305436611, 0.0733107477426529, -0.0431...
843ec5f5-62eb-5b31-be97-4cfa64e263d7
/documentation/tutorials-operations/blue-green-deployment/
Blue-Green Cluster Deployment
Step 2: Load Data into the New Cluster > Option A: Load from Your Primary Data Source
documentation
false
Blue-Green Cluster Deployment > Step 2: Load Data into the New Cluster > Option A: Load from Your Primary Data Source If your primary data source is outside of Qdrant (for example, a database, or an object store), you can load the green cluster directly from there using your own ingestion script. This is often a natur...
[ 0.046819690614938736, -0.07039865106344223, -0.02623499184846878, -0.05531160160899162, 0.10124173760414124, 0.0015627610264346004, 0.03785581886768341, 0.05154965817928314, 0.025915751233696938, -0.04001148045063019, 0.03257078304886818, 0.02623092196881771, 0.024794230237603188, -0.01234...
8451a8de-af66-5c98-b734-363255f2b21b
/documentation/deploy-intro/
Deploy Qdrant
Security & Troubleshooting
documentation
false
Deploy Qdrant > Security & Troubleshooting - [Security](/documentation/security/index.md) — API keys, JWT-based collection-scoped access control, TLS encryption, and network binding. - [Troubleshooting](/documentation/common-errors/index.md) — Diagnose common runtime errors: open-file limits, filesystem incompatibilit...
[ 0.05434809625148773, -0.03813564404845238, -0.050082698464393616, -0.06241156905889511, 0.03377065435051918, -0.03518892452120781, 0.05907245725393295, 0.02207355946302414, 0.08999743312597275, -0.006133774295449257, 0.038349222391843796, -0.004922974389046431, 0.05272110924124718, -0.0358...
845d4290-50b2-50a8-90d9-cdd58df6081b
/documentation/manage-data/indexing/
Indexing
Full-Text Index > Phrase Matching
documentation
true
Indexing > Full-Text Index > Phrase Matching Phrase matching in Qdrant allows you to find documents or points where a specific sequence of words appears together, in the same order, within a text payload field. This is useful when you want to match phrases rather than individual words scattered throughout the text. F...
[ 0.04861539229750633, -0.05811262130737305, -0.03642703592777252, -0.049432478845119476, 0.08590440452098846, -0.03250499069690704, 0.053659383207559586, 0.027646448463201523, 0.04422770068049431, -0.03445214033126831, 0.021638626232743263, 0.009459502063691616, 0.0451744869351387, -0.02154...
845f9b44-431e-5263-9f64-28e1b1e1bf14
/documentation/tutorials-operations/data-streaming-kafka-qdrant/
Kafka Streaming into Qdrant
Kafka
documentation
false
Kafka Streaming into Qdrant > Kafka At the heart of this architecture lies **Kafka**, a distributed event streaming platform capable of handling trillions of events a day. Kafka acts as a central hub where data from various sources can be ingested, processed, and distributed to various downstream systems. Its fault-to...
[ 0.012174716219305992, -0.005333262495696545, -0.07064967602491379, -0.048061076551675797, 0.06672865897417068, -0.020818917080760002, 0.05586809664964676, 0.043852176517248154, 0.04434182122349739, -0.014467138797044754, 0.025882868096232414, -0.013283693231642246, 0.04441022500395775, -0....
846deaf2-4557-5f94-9200-b704432564fe
/documentation/tutorials-develop/search-as-you-type/
Semantic Search As You Type
Step 1: Choose an Embedding Model
documentation
true
Semantic Search As You Type > Step 1: Choose an Embedding Model Clusters on Qdrant Managed Cloud can embed text on the server with [Cloud Inference](/documentation/inference/cloud-inference/). On a self-hosted deployment, Hybrid Cloud, or Private Cloud, Qdrant does not embed text, so you run an embedding model yoursel...
[ 0.06098441779613495, -0.03314914181828499, -0.047009315341711044, -0.0397333949804306, 0.07819875329732895, -0.035336337983608246, 0.09180911630392075, 0.02311132475733757, 0.014990591444075108, -0.009256916120648384, 0.03142736107110977, 0.007261037360876799, 0.04073372855782509, -0.05581...
84a79acb-203b-5fbe-a738-69f3c3afdc3d
/course/essentials/day-5/universal-query-demo/
Demo: Universal Query for Hybrid Retrieval
Prepare and Ingest Research Paper Data
course
true
Demo: Universal Query for Hybrid Retrieval > Prepare and Ingest Research Paper Data Now that our collection is configured with vectors and payload indexes, let's define a few sample research papers: ```python sample_data = [ { "title": "Zero-Shot Retrieval for Scalable Visual Search in a Two-Sided Marketp...
[ 0.04055001586675644, -0.0017570547061040998, -0.05015508458018303, -0.04855896905064583, 0.09572747349739075, -0.0431380569934845, 0.06550537049770355, 0.07699055969715118, 0.07272028923034668, -0.045455433428287506, 0.04665927588939667, 0.03686175495386124, 0.05068421736359596, -0.0158370...
84aa4d61-9c8a-5ae9-83d2-be3ba19f64bf
/documentation/manage-data/points/
Points
Upload Points > Inference
documentation
true
Points > Upload Points > Inference Instead of providing vectors explicitly, Qdrant can also generate vectors using a process called [inference](/documentation/inference/). Inference is the process of creating vector embeddings from text, images, or other data types using a machine learning model. You can use inferenc...
[ 0.05409716069698334, -0.04623592644929886, -0.024311242625117302, -0.07998036593198776, 0.05232871323823929, -0.013361191377043724, 0.12699009478092194, 0.06781312823295593, 0.06723317503929138, -0.01484740525484085, 0.03567516803741455, 0.013372045010328293, 0.051319196820259094, -0.02287...
84b03f5b-5a38-5909-8ae7-2abf17406ed0
/course/multi-vector-search/module-1/use-cases-multi-vector/
Use Cases for Multi-Vector Search
course
false
Use Cases for Multi-Vector Search Module 1 **When is the added complexity of multi-vector search actually worth it?** Multi-vector representations require more storage, more computation, and more careful implementation than simple single-vector embeddings. So why bother? The answer comes down to one core capability:...
[ 0.04045841470360756, -0.059104908257722855, -0.03235136717557907, -0.042066119611263275, 0.0705065205693245, -0.045683830976486206, 0.04128250107169151, 0.02967832237482071, 0.07150136679410934, -0.03343966230750084, 0.02539789490401745, 0.006587809417396784, 0.08343574404716492, -0.028206...
84b2f3c9-fbdf-5620-a903-98a4b842aa53
/documentation/manage-data/quantization/
Quantization
Binary Quantization > Rescoring and Oversampling Results
documentation
false
Quantization > Binary Quantization > Rescoring and Oversampling Results A 2024 sweep over the OpenAI `text-embedding-3` models shows what rescoring and oversampling do to binary quantization in practice. The numbers below come from that run, with the setup that produced them, so you can judge how far they carry to you...
[ 0.06330902129411697, -0.026513516902923584, -0.019085746258497238, -0.07216064631938934, 0.05865512415766716, -0.012289533391594887, 0.05695108324289322, 0.02141532674431801, 0.0681704431772232, -0.03689391911029816, 0.032319437712430954, 0.01645210199058056, 0.08062716573476791, -0.016134...
84c884bd-cdaa-5d19-b462-92d646d79389
/documentation/hybrid-cloud/networking-logging-monitoring/
Networking, Logging & Monitoring
Monitoring
documentation
false
Networking, Logging & Monitoring > Monitoring The Qdrant Cloud console gives you access to basic metrics about CPU, memory and disk usage of your Qdrant clusters. If you want to integrate Qdrant metrics into your own monitoring system, configure it to scrape the following endpoints, which provide metrics in a Prometh...
[ 0.02578471601009369, -0.03249833732843399, -0.021248430013656616, 0.0051361750811338425, 0.06227627769112587, -0.04794684797525406, 0.07386881113052368, 0.06477990001440048, 0.06604526937007904, -0.05486537888646126, 0.03271130472421646, -0.013622701168060303, 0.0348367877304554, -0.011725...
84e62932-8ffa-5eb8-8717-8a34fff99db2
/course/essentials/day-5/colbert-multivectors/
Multivectors for Late Interaction Models
course
false
Multivectors for Late Interaction Models Day 5 Many embedding models represent data as a single vector. Transformer-based encoders achieve this by pooling the per-token vector matrix from the final layer into a single vector. That works great for most cases. But when your documents get more complex, cover multiple to...
[ 0.02988450601696968, -0.04268046095967293, -0.045799579471349716, -0.04176458343863487, 0.015664655715227127, -0.020885268226265907, 0.10398200154304504, 0.04960324242711067, 0.054902348667383194, 0.00370545475743711, 0.004183600191026926, 0.015726014971733093, 0.030295787379145622, -0.051...
84ee8933-8c8d-5a52-900b-9c39c38415a6
/documentation/frameworks/cheshire-cat/
Cheshire Cat
documentation
true
Cheshire Cat [Cheshire Cat](https://cheshirecat.ai/) is an open-source framework that allows you to develop intelligent agents on top of many Large Language Models (LLM). You can develop your custom AI architecture to assist you in a wide range of tasks. ![Cheshire cat](/documentation/frameworks/cheshire-cat/cat.jpg)...
[ 0.06121489405632019, -0.05054408311843872, -0.037927061319351196, -0.014624122530221939, 0.05762527510523796, -0.048855558037757874, 0.05547897517681122, 0.03518771007657051, 0.05157962813973427, -0.038769304752349854, 0.03186439722776413, -0.014399688690900803, -0.006195473950356245, -0.0...
850ed733-7348-5a05-bae8-fb24fbadf727
/documentation/embeddings/mistral/
Mistral
Models
documentation
false
Mistral > Models | Model | Dimensions | Reducible | Max Tokens per Input | |---|---|---|---| | `mistral-embed` | 1024 | No | 8192 | | `codestral-embed` | 1536 (up to 3072) | Yes | 8192 | `mistral-embed` is for general text. `codestral-embed` is for code retrieval. To pin a version, use `mistral-embed-2312` or `codest...
[ 0.05432455986738205, 0.00013825877977069467, -0.01678873598575592, -0.00522981770336628, 0.04677947238087654, -0.08401231467723846, 0.05034039169549942, 0.0789494514465332, 0.04793265834450722, -0.06052999570965767, 0.03036932647228241, -0.010998652316629887, 0.057858437299728394, -0.04220...
85453a3c-c279-5771-9e12-1ceb2d34d34a
/articles/modern-sparse-neural-retrieval/
Modern Sparse Neural Retrieval: From DeepCT to SPLADE++
Did we Solve the Vocabulary Mismatch Yet? > State-of-the-Art of Modern Sparse Neural Retrieval
article
false
Modern Sparse Neural Retrieval: From DeepCT to SPLADE++ > Did we Solve the Vocabulary Mismatch Yet? > State-of-the-Art of Modern Sparse Neural Retrieval The authors of the [`Sparse Lexical and Expansion Model (SPLADE)`](https://arxiv.org/pdf/2109.10086) family of models added dense model training tricks to the intern...
[ 0.05962473899126053, -0.029447805136442184, -0.054602012038230896, -0.02642209827899933, 0.058970291167497635, -0.036320000886917114, 0.07795817404985428, 0.014274253509938717, 0.07036711275577545, -0.023505156859755516, 0.033147864043712616, 0.015290457755327225, 0.04706534743309021, -0.0...
85574b30-1fdd-5953-a71a-d5b1d55bca0f
/documentation/faq/qdrant-fundamentals/
Qdrant Fundamentals
Vectors > Why collection info shows inaccurate number of points?
documentation
false
Qdrant Fundamentals > Vectors > Why collection info shows inaccurate number of points? Collection info API in Qdrant returns an approximate number of points in the collection. If you need an exact number, you can use the [count](/documentation/manage-data/points/#counting-points) API.
[ 0.03324125334620476, -0.04271210357546806, -0.0018576827133074403, -0.06200031191110611, 0.04313793033361435, -0.02545454353094101, 0.06387680023908615, 0.09669677168130875, 0.09486232697963715, -0.02951129898428917, 0.041315507143735886, 0.049933046102523804, 0.05374986305832863, -0.02766...
85589583-e0fd-584a-937a-b6db62f8e6c6
/course/multi-vector-search/module-1/multi-vector-in-qdrant/
Multi-Vector Embeddings in Qdrant
Creating a Multi-Vector Collection
course
true
Multi-Vector Embeddings in Qdrant > Creating a Multi-Vector Collection Setting up a collection for multi-vector search requires specific configuration to enable late interaction and MaxSim distance calculation. Here's how to create a collection configured for ColBERT embeddings: ```python from qdrant_client import Qd...
[ 0.04344501346349716, -0.039113640785217285, -0.03812970221042633, -0.019889013841748238, 0.0656207948923111, -0.04121693596243858, 0.053520966321229935, 0.06945592910051346, 0.05689002200961113, -0.04328208044171333, 0.043704692274332047, 0.026029890403151512, 0.06646442413330078, -0.04051...
85831c83-5c94-56a4-9243-29a30e5001e6
/documentation/manage-data/points/
Points
Upload Points
documentation
true
Points > Upload Points To optimize performance, Qdrant supports batch loading of points. This means you can load several points into the service in one API call. Batching allows you to minimize the overhead of creating a network connection. The Qdrant API supports two ways of creating batches: record-oriented and col...
[ -0.018487347289919853, -0.04360979422926903, -0.04316435381770134, -0.08548358827829361, 0.06325854361057281, -0.04906676337122917, 0.07225801050662994, 0.07520590722560883, 0.0874544084072113, -0.014684094116091728, 0.042433448135852814, -0.004336643032729626, 0.030117450281977654, -0.003...
8585632a-2964-5f61-8ac5-cc2fd0e07921
/documentation/cloud-rbac/permission-reference/
Permission Reference
Profile
documentation
false
Permission Reference > Profile Permissions for accessing personal profile information. | Permission | Description | |------------|------------| | `read:profile` | View the user’s own profile information. **(Assigned to all users by default)** | ---
[ 0.005074080545455217, -0.02155965380370617, -0.036205340176820755, -0.082585908472538, 0.05353944003582001, -0.031297434121370316, 0.019713453948497772, 0.08314535021781921, 0.08046106994152069, -0.011345872655510902, 0.025048408657312393, -0.0019886938389390707, 0.10046936571598053, -0.03...
85856a7f-74b9-5a27-b006-90bc739dc298
/articles/agentic-rag/
What is Agentic RAG? Building Agents with Qdrant
Which Framework is Best? > CrewAI
article
false
What is Agentic RAG? Building Agents with Qdrant > Which Framework is Best? > CrewAI *Disclaimer: QdrantStorage is not a part of the CrewAI framework, but it's taken from the Qdrant documentation on [how to integrate Qdrant with CrewAI](https://qdrant.tech/documentation/frameworks/crewai/).* Although it's not a techn...
[ 0.00823263730853796, -0.0032534447964280844, -0.03052547015249729, -0.035970307886600494, 0.059875715523958206, -0.077564537525177, 0.05053384229540825, 0.025541560724377632, 0.061489079147577286, -0.037018291652202606, 0.033116552978754044, -0.022730587050318718, 0.06435459107160568, 0.00...
8599c77a-a4fd-5c15-8211-5f4e625ece3e
/documentation/tutorials-build-essentials/agentic-rag-crewai-zoom/
Agentic RAG with CrewAI
What You'll Build > How Does It Work?
documentation
false
Agentic RAG with CrewAI > What You'll Build > How Does It Work? When you interact with the system, here's what happens behind the scenes: First the user submits a query to the system. In this example, we want to find out the average length of Marketing meetings. Since one of the data points from the meetings is the d...
[ 0.013617288321256638, -0.048902079463005066, -0.042816974222660065, 0.0085835549980402, 0.07484807819128036, -0.039671529084444046, 0.0725267082452774, 0.05740363150835037, 0.09302651882171631, -0.03399335965514183, 0.030888933688402176, 0.007800841238349676, 0.06876059621572495, -0.028147...
85d63e7b-79c7-5fc0-89ba-92c04384f421
/documentation/manage-data/payload/
Payload
Payload types > Geo
documentation
true
Payload > Payload types > Geo `geo` is used to represent geographical coordinates. Example of single and multiple `geo` values: ```json { "location": { "lon": 52.5200, "lat": 13.4050 }, "cities": [ { "lon": 51.5072, "lat": 0.1276 }, { ...
[ 0.009583745151758194, -0.022195477038621902, -0.017965316772460938, -0.07330255210399628, 0.04274817928671837, -0.02954256162047386, 0.05803189054131508, 0.059561587870121, 0.07140113413333893, -0.046978577971458435, 0.015649931505322456, -0.01189594715833664, 0.06201336532831192, 0.026963...
85de6211-c139-524f-b3c4-8dd191ee12d8
/documentation/tutorials-build-essentials/video-anomaly-edge-part-2/
Video Anomaly Detection Part 2: Edge-to-Cloud Pipeline
Two-Shard Edge Architecture > The Solution: Mutable + Immutable Shards
documentation
true
Video Anomaly Detection Part 2: Edge-to-Cloud Pipeline > Two-Shard Edge Architecture > The Solution: Mutable + Immutable Shards ![Edge shards: mutable shard for live writes with brute-force search, immutable shard with HNSW index synced from cloud, merged and ranked for top-k results](/articles_data/video-anomaly-edge...
[ 0.034821171313524246, -0.04024239256978035, -0.09413544088602066, -0.04927336424589157, 0.048524003475904465, -0.03470557928085327, 0.07681916654109955, 0.03277479112148285, 0.08077222853899002, -0.055649206042289734, 0.030602028593420982, 0.03561025485396385, 0.012114285491406918, -0.0092...
85e893c4-4c1b-5f04-97a0-36af52c4ce71
/articles/io_uring/
Faster Disk I/O for Vector Search Using io_uring
Best Practices
article
false
Faster Disk I/O for Vector Search Using io_uring > Best Practices If your on-disk collection's query performance is of sufficiently high priority to you, enable io\_uring to greatly reduce operating system overhead from disk IO. On the other hand, if your collections are in memory only, activating it will be ineffecti...
[ 0.03217843174934387, -0.07681122422218323, -0.037536099553108215, -0.009035156108438969, 0.07256744056940079, -0.029131051152944565, 0.023487191647291183, 0.018210869282484055, 0.09728536009788513, -0.03525887057185173, -0.00036532984813675284, 0.004710639826953411, 0.03716835007071495, 0....
85f94b47-f491-54ca-b2d8-c2e0482e850f
/documentation/tutorials-build-essentials/agentic-rag-crewai-zoom/
Agentic RAG with CrewAI
Setup
documentation
true
Agentic RAG with CrewAI > Setup 1. **Clone the Repository**: ```bash git clone https://github.com/qdrant/examples.git cd agentic_rag_zoom_crewai ``` 2. **Create and Activate a Python Virtual Environment with Python 3.10 for compatibility**: ```bash python3.10 -m venv venv source venv/bin/activate # Windows: venv\S...
[ 0.03493238985538483, -0.017018374055624008, -0.027315670624375343, -0.0405288003385067, 0.09688814729452133, -0.04775839298963547, 0.008917094208300114, 0.04367454722523689, 0.061777904629707336, -0.04076020047068596, 0.016704630106687546, -0.005644409451633692, 0.07241082936525345, -0.028...
8600b6e1-834b-5c40-be40-d4588f3bb194
/documentation/production-operations/vector-search-in-production/
Vector Search in Production: Scaling, HA & Tuning Guide
5. Tips for Proper Database Administration > Setting up Access Controls in Production
documentation
false
Vector Search in Production: Scaling, HA & Tuning Guide > 5. Tips for Proper Database Administration > Setting up Access Controls in Production **Use Case:** A large enterprise needed to implement access controls where "team A is able to access only collection A, B and C but not collection D" in their Qdrant database....
[ 0.016188541427254677, -0.02147608809173107, -0.06636910885572433, -0.043995510786771774, 0.04759442061185837, -0.03520239144563675, 0.03615497052669525, 0.03612752631306648, 0.08189152181148529, -0.024934353306889534, 0.03228054195642471, 0.009759779088199139, 0.08291257172822952, -0.06001...
8624d0e2-6ce8-5aa7-9802-16e2fdc2046e
/documentation/tutorials-search-engineering/sparse-embeddings-ecommerce-part-4/
Fine-Tuning Sparse Embeddings for E-Commerce Search | Part 4: Specialization vs Generalization
Multi-Domain Training
documentation
false
Fine-Tuning Sparse Embeddings for E-Commerce Search | Part 4: Specialization vs Generalization > Multi-Domain Training To address the generalization problem, we trained a **multi-domain SPLADE model** on combined data from ESCI, WANDS, and Home Depot: roughly 50K training pairs from each dataset, 150K total. The hypo...
[ 0.05174176022410393, 0.0017234302358701825, -0.07043551653623581, -0.004026390146464109, 0.03371680527925491, -0.058152731508016586, 0.058701448142528534, 0.009197985753417015, 0.07754211127758026, -0.02278076484799385, 0.0075390818528831005, 0.038709744811058044, 0.059667084366083145, -0....
863650ca-d0b7-5206-9751-69f792df38d3
/articles/late-interaction-models/
Late Interaction Retrieval with Dense Token Embeddings
Using Output Token Embeddings for Multi-Vector Representations > Experimental Findings
article
false
Late Interaction Retrieval with Dense Token Embeddings > Using Output Token Embeddings for Multi-Vector Representations > Experimental Findings Even the simple `all-MiniLM-L6-v2` model can be applied in a late interaction model fashion, resulting in a positive impact on retrieval quality. However, the best results wer...
[ 0.06512770056724548, 0.002010819735005498, -0.03265959396958351, -0.04173465445637703, 0.08325446397066116, -0.05659330263733864, 0.04633471742272377, 0.04377416893839836, 0.06278402358293533, -0.019149068742990494, 0.0376579612493515, 0.018175872042775154, 0.027629146352410316, -0.0352195...
863c067f-8c8b-5313-96fa-15575ac563a9
/documentation/search-tuning/before-tuning-a-qdrant-collection/
What to Check Before Tuning a Qdrant Collection
Silent Settings Can Break Quality > Hybrid Search
documentation
false
What to Check Before Tuning a Qdrant Collection > Silent Settings Can Break Quality > Hybrid Search Fusion placement matters on sharded collections. `score_threshold` is a risk at any scale when a request moves from single-vector retrieval to fusion. **[Fusion placement](/documentation/search/hybrid-queries/)** At th...
[ 0.05590398982167244, -0.039684318006038666, -0.05179251357913017, -0.05804852396249771, 0.027682477608323097, -0.028102535754442215, 0.05656873434782028, 0.04398135468363762, 0.0894056186079979, -0.04241061210632324, 0.030794154852628708, -0.0016941115027293563, 0.035950254648923874, 0.014...
86533405-b3f1-538d-bc93-e6b034522b35
/documentation/private-cloud/changelog/
Changelog
1.9.6 (2026-02-19)
documentation
false
Changelog > 1.9.6 (2026-02-19) | Component | Version | |-------------------------|---------| | qdrant-kubernetes-api | v1.23.0 | | operator | 2.15.0 | | qdrant-cluster-manager | v0.3.17 | | qdrant-cluster-exporter | 1.7.6 | Latest validated Qdrant version: 1.17.0 * Support for Qdra...
[ 0.0484338104724884, -0.032485801726579666, -0.0553583949804306, -0.03633365035057068, 0.04837002977728844, -0.047675970941782, 0.05638827383518219, 0.03506311774253845, 0.053881824016571045, -0.028519462794065475, 0.016297077760100365, 0.03465352579951286, 0.058472223579883575, -0.06246710...
86541887-1935-5e8a-8532-25f35fa48ad3
/documentation/tutorials-operations/large-scale-search/
Large-Scale Search
Uploading and Indexing
documentation
false
Large-Scale Search > Uploading and Indexing In our experiment, the upload process was going at 5000 points per second. The indexation process was going in parallel with the upload and was happening at the rate of approximately 4000 points per second.
[ -0.01327433716505766, -0.017222465947270393, -0.032116733491420746, -0.024547230452299118, 0.08892180770635605, -0.016084300354123116, 0.02800426073372364, 0.07751626521348953, 0.12883809208869934, -0.021944759413599968, 0.02546321228146553, 0.010732260532677174, 0.00668124295771122, -0.00...
8676b514-9f2b-571d-8eda-c57a1d5c3df1
/documentation/examples/rag-chatbot-red-hat-openshift-haystack/
Private Chatbot for Interactive Learning
documentation
false
Private Chatbot for Interactive Learning | Time: 120 min | Level: Advanced | | | --- | ----------- | ----------- |----------- | With chatbots, companies can scale their training programs to accommodate a large workforce, delivering consistent and standardized learning experiences across departments, locations, and t...
[ -0.011255481280386448, 0.003492962336167693, -0.05115240439772606, -0.02105036936700344, 0.07197555154561996, -0.031713999807834625, 0.029614293947815895, 0.0592646524310112, 0.044917333871126175, -0.039692655205726624, 0.00026490268646739423, 0.02707958221435547, 0.0563836432993412, -0.06...
8676ffde-565b-59eb-8aeb-969bb60f6645
/documentation/datasets/
Practice Datasets
Arxiv.org > Arxiv.org abstracts
documentation
true
Practice Datasets > Arxiv.org > Arxiv.org abstracts This dataset contains embeddings generated from the paper abstracts. Each vector has a payload with the abstract used to create it, along with the DOI (Digital Object Identifier). ```json { "abstract": "Recently Cole and Gkatzelis gave the first constant factor ...
[ 0.0637265145778656, -0.0013769001234322786, -0.03710738942027092, -0.005037460010498762, 0.07339180260896683, -0.02268695831298828, 0.058692898601293564, 0.09368424862623215, 0.02803179994225502, -0.02803969383239746, 0.036340948194265366, 0.039448656141757965, 0.034237414598464966, -0.036...
8687939a-4851-5c40-bdf7-69c5a9c90036
/documentation/frameworks/mastra/
Mastra
Constructor Options
documentation
false
Mastra > Constructor Options | Name | Type | Description | |--------|-----------|-------------------------------------------------------------------------------------------------------| | `url` | `string` | REST URL of t...
[ 0.023951629176735878, -0.04670623689889908, -0.05718277394771576, -0.011972963809967041, 0.025393983349204063, -0.07342825084924698, 0.09992921352386475, 0.021903472021222115, 0.05102475360035896, -0.04349946230649948, 0.008317013271152973, 0.01789354719221592, 0.03308902680873871, -0.0369...
86b351e8-1ae9-5288-83ec-922fb059c862
/documentation/data-synchronization/with-postgres/
Keeping Postgres and Qdrant in Sync
Tier 3: Change Data Capture with Debezium + Redpanda > Redpanda vs. Kafka
documentation
false
Keeping Postgres and Qdrant in Sync > Tier 3: Change Data Capture with Debezium + Redpanda > Redpanda vs. Kafka The example uses Redpanda as the event bus. Redpanda is Kafka-wire-protocol-compatible but ships as a single binary with no ZooKeeper dependency — simpler to run. The consumer code is identical whether you p...
[ 0.04150133207440376, -0.021845322102308273, -0.03199434280395508, -0.05755686014890671, 0.05077284574508667, -0.04369421303272247, 0.11919502913951874, 0.03852890804409981, 0.019179055467247963, 0.01646578684449196, 0.02699667029082775, 0.027632012963294983, 0.02032308094203472, -0.0378974...
86b704d5-4ba4-59d4-84c4-bf2a2b933f2f
/course/essentials/day-5/colbert-multivectors/
Multivectors for Late Interaction Models
Collection Configuration: Multivector Setup
course
true
Multivectors for Late Interaction Models > Collection Configuration: Multivector Setup To use ColBERT for retrieval, create a collection with a multivector field configured for late interaction scoring: ```python from qdrant_client import QdrantClient, models import os client = QdrantClient(url=os.getenv("QDRANT_URL...
[ 0.04311560094356537, -0.023891694843769073, -0.018050484359264374, -0.021119514480233192, 0.0670056939125061, -0.06171531230211258, 0.05446511134505272, 0.04931606724858284, 0.03929251804947853, -0.018563920632004738, 0.04934685304760933, 0.02386249229311943, 0.07020677626132965, -0.027734...
86d76864-493e-55b1-b8fb-d165aa9253db
/course/essentials/day-2/collection-tuning-demo/
Demo: HNSW Performance Tuning
Step 10: Create Payload Indexes
course
true
Demo: HNSW Performance Tuning > Step 10: Create Payload Indexes Create a [full‑text index](/documentation/manage-data/indexing/#full-text-index) for faster filtering. ```python # Create a payload index for 'text' so filters use an index, not a scan. client.create_payload_index( collection_name=collection_name, ...
[ 0.015593593940138817, -0.02439776621758938, -0.04250609874725342, -0.04244859516620636, 0.05637944117188454, -0.03644080087542534, 0.016440870240330696, 0.026893319562077522, 0.06039183959364891, -0.021670745685696602, 0.03994828835129738, 0.05364473536610603, 0.05876944214105606, -0.03357...
86f94ca9-38a6-51dd-8a1b-da474d44f2cd
/documentation/migration-guidance/search-quality/
Search Quality
Tier 2: Statistical Sampling (Recommended) > Extending Tier 2: Score Correlation
documentation
true
Search Quality > Tier 2: Statistical Sampling (Recommended) > Extending Tier 2: Score Correlation Beyond recall, check whether the relative ordering of scores is consistent: ```py from scipy import stats as scipy_stats def score_correlation(source_results, qdrant_results): """Compute rank correlation between sou...
[ 0.04038609191775322, -0.04662608355283737, -0.022442860528826714, -0.04347154498100281, 0.03894875571131706, -0.027454407885670662, 0.06824546307325363, 0.0630960538983345, 0.10785416513681412, -0.024606965482234955, 0.02123245969414711, 0.020052360370755196, 0.06553717702627182, 0.0213436...
8729b143-ff1d-5262-904a-e03ef40de3dd
/documentation/tutorials-basics/search-beginners-local/
Semantic Search 101
1. Installation
documentation
true
Semantic Search 101 > 1. Installation You need to process your data so that the search engine can work with it. The [Sentence Transformers](https://www.sbert.net/) framework gives you access to common Large Language Models that turn raw data into embeddings. ```bash pip install -U sentence-transformers ``` Once enco...
[ 0.029902148991823196, -0.06250862777233124, -0.02556844800710678, -0.010836519300937653, 0.016093553975224495, -0.03458095341920853, 0.08215218782424927, 0.024110082536935806, 0.05213085189461708, -0.02713877521455288, 0.03455489128828049, 0.024580590426921844, 0.07049392908811569, -0.0573...
8731dc21-9553-5ef6-9315-8aba6a46f548
/documentation/cloud-user-profile/
User Profile
Delete User
documentation
false
User Profile > Delete User Use **Delete User** on the **Preferences** page to permanently delete your Qdrant user and all associated data. A deleted user cannot be restored. * **No accounts:** you are asked to type your email address to confirm the deletion. * **Owner of one or more accounts:** for each account you o...
[ 0.035158198326826096, -0.046622324734926224, -0.012426961213350296, -0.021011952310800552, 0.0480782613158226, -0.0263542328029871, 0.02091865800321102, 0.042735226452350616, 0.07901835441589355, -0.049878835678100586, 0.028683722019195557, 0.02757066674530506, 0.08953157812356949, -0.0510...
87349291-f350-5d21-891b-6a157b5284dd
/documentation/manage-data/vectors/
Vectors
Datatypes > TurboQuant Datatypes
documentation
false
Vectors > Datatypes > TurboQuant Datatypes TurboQuant datatypes store each vector in a compressed form, using fewer bits per dimension than a 32-bit float. They're built on [TurboQuant](/articles/turboquant-quantization/). Each vector is first rotated mathematically, spreading each dimension's information evenly acros...
[ 0.028187667950987816, -0.03773193061351776, -0.042348649352788925, -0.036987993866205215, 0.01130166370421648, 0.0005706508527509868, 0.06065751239657402, 0.010377072729170322, 0.06526828557252884, -0.040578246116638184, 0.048334382474422455, 0.039954572916030884, 0.07180488109588623, 0.00...
87436a34-e70f-5f08-acdd-d73f08e5d527
/documentation/migrate-to-qdrant/from-turbopuffer/
From Turbopuffer
Gotchas
documentation
false
From Turbopuffer > Gotchas - **Indexes:** The tool does not carry over Turbopuffer's full-text search or filterable settings. Set up equivalent [payload indexes](/documentation/manage-data/indexing/#payload-index) in Qdrant after the migration. ## Next Steps After migration, verify your data arrived correctly with t...
[ 0.024018067866563797, -0.06530362367630005, -0.03577474132180214, -0.02885226160287857, 0.06758705526590347, 0.0197170190513134, 0.04635974392294884, 0.011166476644575596, 0.03933976590633392, -0.008350593969225883, 0.044423311948776245, 0.03747496008872986, 0.036073215305805206, -0.047042...
87516b73-e26e-5255-83bb-d0e1bf232d9c
/documentation/tutorials-search-engineering/multi-representation-search/
Multi-Representation Search
Retrieval > When to Boost, When to Rerank
documentation
false
Multi-Representation Search > Retrieval > When to Boost, When to Rerank Score boosting expresses ranking preferences that aren't captured by retrieval scores alone: recency, source authority, geographic proximity, content type. Use a FormulaQuery whose terms reference payload fields. For time-based decay on a `publish...
[ 0.017980610951781273, -0.021007461473345757, -0.05149541050195694, -0.06544803082942963, 0.07382047921419144, -0.030238866806030273, 0.018555164337158203, 0.00943143293261528, 0.07443739473819733, -0.03418581932783127, 0.04906969144940376, 0.015595783479511738, 0.0762847289443016, 0.012542...
87b63f2f-1095-5515-9e4c-cf2cc6988828
/documentation/examples/rag-chatbot-red-hat-openshift-haystack/
Private Chatbot for Interactive Learning
Implementation > Search pipeline
documentation
true
Private Chatbot for Interactive Learning > Implementation > Search pipeline Our documents are now indexed and ready for search. The next pipeline is a bit simpler, but we still need to define a few components. Let's start again with an empty pipeline: ```python search_pipeline = Pipeline() ``` Our second process tak...
[ 0.020992662757635117, -0.055498018860816956, -0.03050548955798149, -0.02324146404862404, 0.05507943034172058, -0.0263986773788929, 0.03424893319606781, 0.050350386649370193, 0.031305942684412, -0.02235986664891243, 0.01814856007695198, 0.010590218007564545, 0.030955351889133453, -0.0512768...
87d99446-01a8-58c8-8cc6-89f7c1ace2b6
/documentation/manage-data/indexing/
Indexing
Payload Index > Parameterized Index > Per-Element Index for Arrays of Objects
documentation
false
Indexing > Payload Index > Parameterized Index > Per-Element Index for Arrays of Objects * The field must be inside the elements of the array, such as `offers[].price` for the array `offers`. * Only one array level is supported. `nested` can't point to an array inside another array, such as `a[].b[]`. * A per-element ...
[ 0.005297122988849878, -0.003475320991128683, -0.0463719516992569, -0.02360411360859871, 0.04668504744768143, -0.03596944361925125, -0.004450588021427393, 0.03435293585062027, 0.08547652512788773, -0.021289454773068428, 0.0247499942779541, 0.04802564159035683, 0.0712549239397049, 0.00743852...
87e5ae96-bdcc-535c-aa0c-f77920561a7b
/documentation/fastembed/fastembed-colbert/
Working with ColBERT
Embed data
documentation
true
Working with ColBERT > Embed data We will vectorize a toy movie description dataset with ColBERT: Movie description dataset ```python descriptions = ["In 1431, Jeanne d'Arc is placed on trial on charges of heresy. The ecclesiastical jurists attempt to force Jeanne to recant her claims of holy visions.", "A film pro...
[ 0.0486319400370121, -0.015352681279182434, -0.04631110653281212, -0.06990835815668106, 0.0905655175447464, -0.03598082810640335, 0.06812434643507004, 0.07314685732126236, 0.06339742988348007, -0.004199605900794268, 0.03355541452765465, 0.007192447315901518, 0.07158706337213516, -0.05841295...
87f525c6-5a3f-5d5a-b248-6e661578eb6a
/documentation/tutorials-search-engineering/multi-representation-search/
Multi-Representation Search
Retrieval > What to Prefetch
documentation
false
Multi-Representation Search > Retrieval > What to Prefetch A representation only earns its own prefetch if it carries signal independent of the others. The four above are deliberately complementary: - `dense_chunk` carries the body content. - `dense_title` carries the topical naming. The title isn't part of the chunk...
[ 0.028178200125694275, -0.04259003326296806, -0.0404951311647892, -0.05454952269792557, 0.0749148353934288, -0.053933508694171906, 0.05468041077256203, 0.05495426058769226, 0.0436132550239563, -0.0330817848443985, 0.025084450840950012, 0.05224544182419777, 0.03668995201587677, -0.0393438562...
87fb8a17-4123-5a35-ad24-dff249bf99de
/documentation/search-precision/reranking-semantic-search/
Reranking for Better Search
Implementing Vector Search with Reranking > Overview
documentation
false
Reranking for Better Search > Implementing Vector Search with Reranking > Overview A typical search system works in two main stages: Ingestion and Retrieval. Think of ingestion as the process where your data gets prepped and loaded into the system, and retrieval as the part where the magic happens—where your queries p...
[ -0.012240292504429817, -0.08989547193050385, -0.05566061660647392, -0.014555745758116245, 0.08228536695241928, -0.022823743522167206, 0.03693651780486107, 0.04971177875995636, 0.09991346299648285, -0.008018555119633675, 0.01110850553959608, 0.0326988659799099, 0.03690878301858902, -0.03441...
88044e14-c230-53b5-99ba-c10a9376224a
/articles/sparse-vectors/
Understanding SPLADE and Sparse Vectors
Why SPLADE works: term expansion > Interpreting SPLADE
article
false
Understanding SPLADE and Sparse Vectors > Why SPLADE works: term expansion > Interpreting SPLADE A downside of dense vectors is that they are not interpretable, making it difficult to understand why a document is relevant to a query. SPLADE importance estimation can provide insights into the 'why' behind a document's...
[ 0.023723777383565903, -0.05032647028565407, -0.03835815191268921, -0.06397600471973419, 0.009844714775681496, -0.04208055138587952, 0.059815868735313416, 0.06878883391618729, 0.0491517074406147, -0.013278326019644737, 0.02193647436797619, 0.022501777857542038, 0.022899610921740532, 0.03775...
88119ea7-7ce0-530e-b480-52103e1d78f7
/documentation/faq/database-optimization/
Database Optimization
How do I reduce memory usage?
documentation
false
Database Optimization > How do I reduce memory usage? The primary source of memory usage is vector data. There are several ways to address that: - Configure [Quantization](/documentation/manage-data/quantization/) to reduce the memory usage of vectors. - Configure on-disk vector storage The choice of the approach de...
[ 0.03771669417619705, -0.04508604109287262, -0.018575001507997513, -0.04189049080014229, 0.06345382332801819, 0.0011314812581986189, 0.0427376888692379, 0.04109170287847519, 0.09264019131660461, -0.016866421326994896, 0.010173637419939041, 0.005302026867866516, 0.038542404770851135, -0.0438...
88581bf9-951d-5791-9444-37de0ada424d
/articles/filtered-vector-search-acorn/
Filtered Vector Search: What ACORN Fixes, and What Fixes ACORN
article
false
Filtered Vector Search: What ACORN Fixes, and What Fixes ACORN Filtered vector search breaks when metadata filters turn a healthy nearest-neighbor graph into scattered islands. HNSW's `m` parameter controls how many links each point gets. At Qdrant's default `m=16`, the one-million-point collection benchmarked in this...
[ 0.058381956070661545, -0.06920432299375534, -0.06751827895641327, -0.0015662951627746224, 0.06438270956277847, -0.018283970654010773, 0.047569915652275085, 0.04597284272313118, 0.0660148486495018, 0.004347929265350103, 0.04222479090094566, 0.04555957391858101, 0.03991362079977989, -0.03285...
88660864-0b63-5ac2-824e-04cfe3b98fab
/documentation/tutorials-develop/serverless-semantic-search/
Serverless Semantic Search
Discussion
documentation
false
Serverless Semantic Search > Discussion Lambda starts your function when its URL is called, so no compute stays on hand while it is idle. The first call after an idle period needs extra time to start a new execution environment, and later calls reuse it. Every call also includes the latency of the embedding provider a...
[ 0.025452736765146255, -0.030643818899989128, -0.01825290359556675, -0.02863486111164093, 0.04535656422376633, -0.06218107044696808, 0.07802137732505798, 0.02738453447818756, 0.07330821454524994, -0.028425758704543114, 0.04417409375309944, -0.012091007083654404, 0.011247679591178894, -0.028...
88697973-e7db-596e-875e-07feba01b64d
/documentation/tutorials-search-engineering/pdf-retrieval-at-scale/
Multivector Document Retrieval
documentation
false
Multivector Document Retrieval ![scaling-pdf-retrieval-qdrant](/documentation/tutorials/pdf-retrieval-at-scale/image1.png) | Time: 30 min | Level: Intermediate |Output: [GitHub](https://github.com/qdrant/examples/blob/master/pdf-retrieval-at-scale/ColPali_ColQwen2_Tutorial.ipynb)|[![Open In Colab](https://colab.resea...
[ 0.020158149302005768, -0.04158031940460205, -0.055306628346443176, -0.058174435049295425, 0.09309324622154236, -0.02358095347881317, 0.026788176968693733, 0.026216955855488777, 0.08509328216314316, -0.023528924211859703, 0.009373044595122337, -0.014088096097111702, 0.05051251873373985, -0....
8877a4b1-a50e-549a-b344-ee607f5f07b5
/documentation/cloud-security/
Cloud Security
Compliance and Certifications
documentation
false
Cloud Security > Compliance and Certifications Qdrant is committed to maintaining high standards of security and compliance. We are both SOC2 Type 2 and HIPAA certified, ensuring that our systems and processes meet rigorous security criteria. You can find our compliance reports in our [Trust Center](https://qdrant.to/...
[ 0.028123632073402405, -0.0035548375453799963, -0.05249280855059624, -0.06992006301879883, -0.00729892123490572, -0.047036100178956985, 0.10353437066078186, 0.08174450695514679, 0.09398376941680908, -0.018149496987462044, 0.022340109571814537, 0.015547452494502068, 0.07315754145383835, -0.0...
887b6087-65ca-5fcb-bdb2-c254b60cf74b
/documentation/tutorials-develop/async-api/
Async API
When to use async API
documentation
false
Async API > When to use async API There is no need to use async API if the application you are writing will never support multiple users at once (e.g it is a script that runs once per day). However, if you are writing a web service that multiple users will use simultaneously, you shouldn't be blocking the threads of ...
[ -0.01104031689465046, 0.0019299433333799243, -0.034421537071466446, -0.034378912299871445, 0.020599903538823128, -0.060075271874666214, 0.02482672594487667, 0.01721172221004963, 0.07276073843240738, -0.014472653158009052, 0.008745342493057251, 0.033253468573093414, 0.07575883716344833, -0....
889028fc-2c89-530a-a1ea-5c10454d5755
/documentation/embeddings/nvidia/
NVIDIA
Embed Images
documentation
true
NVIDIA > Embed Images `nvidia/llama-nemotron-embed-vl-1b-v2` embeds documents as images, text, or both, and retrieves them with text queries. Pass each image as a PNG or JPEG data URL with `modality` set to `image`. Images are always documents, so they need `input_type` set to `passage`. This model's vectors don't sha...
[ 0.03512303903698921, -0.0720079243183136, -0.048000678420066833, -0.08431823551654816, 0.07231710106134415, -0.0225857924669981, 0.06063457578420639, 0.06366584450006485, 0.024484561756253242, -0.03319163993000984, 0.023705361410975456, 0.01003884244710207, 0.03045377880334854, -0.00896242...
88bf6992-3aa3-5491-ae34-e0e11342e0cd
/documentation/
Documentation
API Reference
documentation
false
Documentation > API Reference - [Qdrant API Reference](https://api.qdrant.tech/api-reference) — Full REST API reference for all Qdrant operations: collections, points, search, indexing, cluster management, and more.
[ 0.006066028960049152, -0.025799129158258438, -0.029183657839894295, -0.057586416602134705, 0.04106416925787926, -0.03216664120554924, 0.057433873414993286, 0.07874134182929993, 0.08092595636844635, -0.03842414543032646, 0.0548403337597847, 0.041057318449020386, 0.03762557730078697, -0.0422...
88d1aa99-fc2f-571e-a294-3fc112827bd8
/documentation/capacity-planning/
Capacity Planning
Choosing Disk over RAM > Storage-Focused Configuration
documentation
false
Capacity Planning > Choosing Disk over RAM > Storage-Focused Configuration If your priority is to handle large volumes of vectors with average search latency, it's recommended to move vectors to the [`cold` memory tier](/documentation/ops-configuration/memory-tiers/). In this setup, vectors are stored on disk in memor...
[ 0.001559810247272253, -0.05683520808815956, -0.04260731115937233, -0.05853860825300217, 0.04460066556930542, -0.03880096599459648, 0.04357924684882164, 0.011956814676523209, 0.09417882561683655, -0.01726650819182396, -0.014036502689123154, 0.017521925270557404, 0.0543021634221077, -0.01956...
88d1da45-567f-5c41-804e-9e86ff904ee6
/documentation/fastembed/fastembed-postprocessing/
Multi-Vector Postprocessing
documentation
false
Multi-Vector Postprocessing FastEmbed's postprocessing module provides techniques for transforming and optimizing embeddings after generation. These postprocessing methods can improve search performance, reduce storage requirements, or adapt embeddings for specific use cases. Currently, the postprocessing module in...
[ 0.07202572375535965, -0.04792908579111099, -0.004584179259836674, -0.07491382956504822, 0.08759389817714691, -0.04615610092878342, 0.059950508177280426, -0.018106359988451004, 0.06309443712234497, 0.00653134286403656, 0.012041522189974785, 0.07332559674978256, 0.010109666734933853, -0.0223...
88d301bd-4b26-5178-9eb4-47c62a748f63
/course/multi-vector-search/module-1/maxsim-distance/
MaxSim Distance Metric
The MaxSim Formula
course
false
MaxSim Distance Metric > The MaxSim Formula In late interaction, we represent documents and queries as sequences of token vectors. But how do we measure similarity between two sets of vectors? The answer is **MaxSim** (Maximum Similarity), defined mathematically as: $$ \text{MaxSim}(Q, D) = \sum_{i=1}^{|Q|} \max_{j=...
[ 0.05460234358906746, -0.007590497378259897, -0.0396617129445076, -0.024728864431381226, 0.05815036594867706, -0.03485186770558357, 0.05017857998609543, 0.03156425803899765, 0.07416997849941254, -0.04424577206373215, 0.04423407465219498, 0.0043589151464402676, 0.060187049210071564, -0.01555...
88e57f69-ea58-5570-b3f2-1efcca0df1fa
/documentation/private-cloud/api-reference/
API Reference
qdrant.io/v1 > Resource Types > VolumeSnapshotInfo
documentation
false
API Reference > qdrant.io/v1 > Resource Types > VolumeSnapshotInfo _Appears in:_ - [QdrantClusterSnapshotStatus](#qdrantclustersnapshotstatus) | Field | Description | Default | Validation | | --- | --- | --- | --- | | `volumeSnapshotName` _string_ | VolumeSnapshotName is the name of the volume snapshot | | | | `vol...
[ 0.025690579786896706, -0.009967011399567127, 0.005378589499741793, -0.04062037169933319, 0.059240859001874924, -0.0392327643930912, 0.03079180233180523, 0.05563071742653847, 0.06046740338206291, -0.06141987070441246, 0.05364345759153366, 0.02303665690124035, 0.04258285090327263, -0.0238077...