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 |
|---|---|---|---|---|---|---|---|
596e5a28-05f2-5eb4-bb64-68acf676c3a4 | /documentation/edge/ | Qdrant Edge | Using Qdrant Edge | documentation | false | Qdrant Edge > Using Qdrant Edge
| Type | Guide | What you'll learn |
|--------------|----------------------------------------------------------------... | [
0.03834668919444084,
-0.02103794366121292,
-0.05140147730708122,
-0.02805979922413826,
0.09346325695514679,
-0.0522359162569046,
0.08233549445867538,
0.03786090388894081,
0.0774507001042366,
-0.04093843698501587,
0.052834294736385345,
0.023093659430742264,
0.011463678441941738,
-0.02781128... |
59834929-8c89-5d4f-a136-ee648d2ac95e | /documentation/migration-guidance/ | Migration Guidance | Prerequisites | documentation | false | Migration Guidance > Prerequisites
Before starting verification, you need:
- Access to your source system (to capture baselines)
- A Qdrant instance with your migrated data loaded
- Python 3.8+ with the `qdrant-client` library installed (code examples use Python, but the concepts apply to any client)
- The [Qdrant Mi... | [
0.03983303904533386,
-0.07337946444749832,
-0.012021239846944809,
-0.045675672590732574,
0.05176851898431778,
-0.02420896664261818,
0.0502258837223053,
0.04717370495200157,
0.05164625495672226,
-0.034316036850214005,
0.02516627497971058,
0.011638983152806759,
0.08190456032752991,
-0.041899... |
59906f06-e978-51f1-be1a-1634daef44fe | /documentation/ops-configuration/usage-statistics/ | Usage Statistics | How do we anonymize data? | documentation | false | Usage Statistics > How do we anonymize data?
We understand that some users may be concerned about the privacy of their data.
That is why we make an extra effort to ensure your privacy.
There are several different techniques that we use to anonymize the data:
- We use a random UUID to identify instances. This UUID is... | [
0.035253848880529404,
-0.02791363187134266,
-0.041640546172857285,
-0.04237998276948929,
0.06180015578866005,
-0.017939353361725807,
0.01195553783327341,
0.01655854471027851,
0.05587739497423172,
-0.01678989827632904,
0.02999127469956875,
0.03001801110804081,
0.0791131854057312,
-0.0103018... |
59984913-cc37-5d59-8104-68798ec9fdb7 | /course/beginners/module-5/signal-sources-and-embedding-models/ | Signal Sources and Embedding Models | Text: Dense and Sparse | course | true | Signal Sources and Embedding Models > Text: Dense and Sparse
News articles, filings, and transcripts each get two text vectors, a dense one for meaning and a sparse one for exact tokens: the hybrid pairing from Module 3, using the same two models that module used.
```python
from qdrant_client import QdrantClient, mod... | [
0.03391672298312187,
-0.011186087504029274,
-0.06924925744533539,
-0.05985156074166298,
0.07608917355537415,
-0.04928843304514885,
0.05598960816860199,
0.065494105219841,
0.06751932203769684,
-0.030392972752451897,
0.028234736993908882,
0.030608657747507095,
0.026498941704630852,
-0.010721... |
599f0967-000e-52aa-8710-9d8d539734f4 | /documentation/edge/edge-quickstart/ | Quickstart | Query Points | documentation | true | Quickstart > Query Points
To query points in the Edge Shard, use the [`query`](/documentation/edge/edge-api/reading-data/#query) method:
```python
from qdrant_edge import Query, QueryRequest
results = edge_shard.query(
QueryRequest(
query=Query.Nearest([0.2, 0.1, 0.9, 0.7], using=VECTOR_NAME),
li... | [
0.03616533428430557,
-0.04305831715464592,
-0.00931908655911684,
-0.02386869676411152,
0.05986400321125984,
-0.05143795534968376,
0.029569393023848534,
0.03351542353630066,
0.09439778327941895,
-0.012288670055568218,
0.06247543543577194,
0.010206189006567001,
0.021560756489634514,
-0.00852... |
59ce3aa2-5121-5f75-9fdc-51735e151a88 | /documentation/private-cloud/api-reference/ | API Reference | qdrant.io/v1 > Resource Types > Pause | documentation | false | API Reference > qdrant.io/v1 > Resource Types > Pause
_Appears in:_
- [QdrantClusterSpec](#qdrantclusterspec)
| Field | Description | Default | Validation |
| --- | --- | --- | --- |
| `owner` _string_ | Owner specifies the owner of the pause request. | | |
| `reason` _string_ | Reason specifies the reason for the ... | [
0.02771458588540554,
-0.04267147555947304,
0.005088518839329481,
-0.05474139750003815,
0.05920916423201561,
-0.044797636568546295,
0.06623770296573639,
0.06760752946138382,
0.03466344624757767,
-0.05103094503283501,
0.032831281423568726,
0.005928002297878265,
0.07801193743944168,
-0.036586... |
59d8c3a3-c024-5086-96c0-79ef86ae0226 | /documentation/migration-guidance/data-integrity/ | Data Integrity | Passing Criteria | documentation | false | Data Integrity > Passing Criteria
| Check | Pass | Investigate |
| ----- | ----- | ----- |
| Vector count | Exact match (or within documented tolerance) | Any unexplained difference |
| Dimensions | Exact match | Any mismatch (stop here) |
| Distance metric | Maps correctly to Qdrant equivalent | Any mismatch (stop he... | [
0.06657382845878601,
-0.035654712468385696,
-0.059630244970321655,
-0.04542312026023865,
0.03416145220398903,
-0.0066222865134477615,
0.0562501884996891,
0.014888590201735497,
0.08803761750459671,
-0.05103887990117073,
0.02141251042485237,
0.026029769331216812,
0.06806876510381699,
0.00072... |
59e5dcd8-14af-5712-be21-cb0ceb316975 | /documentation/tutorials-basics/multimodal-search/ | Multimodal Search | Connect to Qdrant | documentation | true | Multimodal Search > Connect to Qdrant
1. **Create a client object for Qdrant, with Cloud Inference enabled**.
You'll use a [Qdrant Cloud Free Tier Cluster](/documentation/cloud/create-cluster/#free-clusters). [Create a free cluster](https://cloud.qdrant.io/), save the associated API key and endpoint URL, and instanti... | [
0.06754238158464432,
-0.04281856119632721,
-0.04464026913046837,
-0.03775293380022049,
0.07538192719221115,
-0.02136080525815487,
0.0819491371512413,
0.02805824764072895,
0.05859561637043953,
-0.031805749982595444,
0.021113982424139977,
-0.0000067337950895307586,
0.07512583583593369,
-0.05... |
59f84d27-7675-505e-a162-9dcee2df2b04 | /documentation/scaling/distributed_deployment/ | Distributed Deployment | Enabling Distributed Mode in Self-Hosted Qdrant | documentation | true | Distributed Deployment > Enabling Distributed Mode in Self-Hosted Qdrant
To enable distributed deployment - enable the cluster mode in the [configuration](/documentation/ops-configuration/configuration/) or using the ENV variable: `QDRANT__CLUSTER__ENABLED=true`.
```yaml
cluster:
# Use `enabled: true` to run Qdrant... | [
0.05813620239496231,
-0.051644254475831985,
-0.06126761808991432,
0.006480750627815723,
0.08224000036716461,
-0.04030284285545349,
0.058232638984918594,
0.08017796277999878,
0.0576145239174366,
0.009989081881940365,
0.017570048570632935,
-0.0045037642121315,
0.0439656525850296,
-0.05700012... |
59fb4e64-3775-57a8-88b6-af5d393f50e0 | /documentation/search-evaluation/pipeline-output-quality/ | Evaluating Pipeline Output Quality | Wiring the RAG Pipeline | documentation | true | Evaluating Pipeline Output Quality > Wiring the RAG Pipeline
The prompt above is a starting point; tune it for your domain: answer style, refusal behavior, whether outside knowledge is allowed, and output format.
**3. Run retrieval and generation.** For each entry, retrieve the top-k chunks, pass them through the gen... | [
0.014781863428652287,
-0.04074444621801376,
-0.020681383088231087,
-0.05108367279171944,
0.07504449039697647,
-0.0218742024153471,
0.0653228685259819,
0.023135321214795113,
0.0397833026945591,
-0.05275064706802368,
-0.003779903519898653,
0.005209964234381914,
0.08698385953903198,
-0.041655... |
5a0db5d0-2b22-5bac-a417-f2f17762acb7 | /documentation/tutorials-search-engineering/using-relevance-feedback/ | Relevance Feedback Retrieval in Qdrant | Customizing for Your Use Case > Training | documentation | true | Relevance Feedback Retrieval in Qdrant > Customizing for Your Use Case > Training
#### Training Data
How well the relevance feedback formula fits your data depends on the amount and **quality** of the training data (how realistic it is for your use case).
In general, we need **a very little amount** of it (50-300 q... | [
0.02247200906276703,
-0.028590206056833267,
-0.05306914076209068,
-0.052913229912519455,
0.05525793135166168,
-0.023782407864928246,
0.03912312537431717,
0.014145292341709137,
0.06664194166660309,
-0.026581179350614548,
0.026416845619678497,
0.0001276724215131253,
0.10592949390411377,
-0.0... |
5a19af15-622a-5d3b-8a00-5f231630f441 | /course/multi-vector-search/module-2/colpali-family/ | ColPali Family Overview | The Impact of Bidirectional Attention | course | false | ColPali Family Overview > The Impact of Bidirectional Attention
Bidirectional attention has emerged as a promising approach for multi-vector representations of multi-modal data.
The choice between **unidirectional** and **bidirectional** attention mechanisms significantly impacts model performance for embedding tasks... | [
0.034561946988105774,
-0.00855585653334856,
-0.03393266722559929,
-0.025000199675559998,
0.05060918629169464,
0.008030363358557224,
0.08197949081659317,
0.06443297863006592,
0.07364797592163086,
-0.0005263723433017731,
0.046682462096214294,
-0.009647266007959843,
0.05619930848479271,
-0.00... |
5a2c8655-e946-5c5e-9f6e-b9899e01c34a | /documentation/ops-monitoring/monitoring/ | Monitoring & Telemetry | Metrics > Node Metrics /metrics | documentation | false | Monitoring & Telemetry > Metrics > Node Metrics /metrics
| Name | Type | Meaning |
| -------------------------------- | ------- | ----------------------------------------------------------------------- |
| cluster_enabled ... | [
0.026100756600499153,
-0.003219643607735634,
-0.06331511586904526,
-0.004132429137825966,
0.04136112332344055,
-0.02971533313393593,
0.0788273885846138,
0.09534327685832977,
0.08445078879594803,
-0.03295329585671425,
0.0443989522755146,
0.0017496455693617463,
0.03185199946165085,
-0.025839... |
5a386d79-0c3b-586b-b39d-5d4e4408e855 | /documentation/cloud-user-profile/ | User Profile | Cookie Consent Preferences | documentation | false | User Profile > Cookie Consent Preferences
The **Cookie Consent Preferences** section lets you control how cookies are used. Cookies are grouped into categories that you can allow or deny individually. Disabling a previously allowed category removes its cookies from your browser. Use **Manage Consent** to review the ca... | [
0.03650970757007599,
0.011949858628213406,
-0.05746222659945488,
-0.0683365985751152,
0.0038251192308962345,
-0.020279670134186745,
0.049903158098459244,
0.03811893239617348,
0.07772604376077652,
-0.045448318123817444,
0.03170047700405121,
0.03308587521314621,
0.08259972929954529,
0.002574... |
5a39e868-ccfb-5f6a-8bfb-c1656b94a03c | /course/essentials/day-6/final-project/ | Final Project: Production-Ready Documentation Search Engine | Optional: Go Further | course | false | Final Project: Production-Ready Documentation Search Engine > Optional: Go Further
* Try different dense models (speed vs quality)
* Rebuild with higher `m` / `ef_construct` if your dataset is large
* Add filters (tags, section type) and measure impact | [
0.033810798078775406,
-0.025008220225572586,
-0.044268909841775894,
-0.023005705326795578,
0.0828995406627655,
-0.02142413705587387,
0.04355364665389061,
0.016355272382497787,
0.06502151489257812,
-0.039608411490917206,
0.037339240312576294,
0.038512345403432846,
0.07159166038036346,
-0.05... |
5a5d7143-e693-541e-aa30-ca5966845538 | /documentation/hybrid-cloud/platform-deployment-options/ | Deployment Platforms | Vultr | documentation | false | Deployment Platforms > Vultr
[Vultr Kubernetes Engine (VKE)](https://www.vultr.com/kubernetes/) is a fully-managed product offering with predictable pricing that makes Kubernetes easy to use. Vultr manages the control plane and worker nodes and provides integration with other managed services such as Load Balancers, B... | [
0.03293401747941971,
-0.03366078808903694,
-0.09323267638683319,
-0.031103024259209633,
0.058181408792734146,
-0.006777339149266481,
0.11216895282268524,
0.0606926754117012,
0.07491693645715714,
-0.023381225764751434,
0.04686931520700455,
0.017096634954214096,
0.01872618868947029,
-0.02731... |
5a7556cf-b1ce-5175-ab07-d783b8b87d49 | /documentation/tutorials-operations/secure-qdrant/ | Secure a Self-Hosted Qdrant Instance | Step 5: Set Up Granular Access API Keys (JWT) | documentation | true | Secure a Self-Hosted Qdrant Instance > Step 5: Set Up Granular Access API Keys (JWT)
The admin and read-only keys apply globally. For finer control, use [granular access API Keys](/documentation/security/#granular-access-api-keys) (JSON Web Tokens, JWT). For example, you can use JWT to provide read-write access to one... | [
0.023045150563120842,
-0.014261141419410706,
-0.03792529180645943,
-0.0319766029715538,
0.05409063771367073,
-0.04814198613166809,
0.05357779935002327,
0.05641492083668709,
0.060386791825294495,
-0.028681481257081032,
0.05235731229186058,
0.04527778923511505,
0.07272306084632874,
-0.028853... |
5a7f471d-d9fd-5079-9e42-0323ee801eba | /documentation/tutorials-build-essentials/video-anomaly-edge-part-3/ | Video Anomaly Detection Part 3: Scoring, Governance, and Deployment | Results > How the scores are computed | documentation | false | Video Anomaly Detection Part 3: Scoring, Governance, and Deployment > Results > How the scores are computed
At the operating point where the false positive rate on normal test clips equals 2 per 360 clips (about 0.56%), recall was 94.2%. If a camera produced 360 normal clips an hour with that same false positive rate,... | [
0.03395498916506767,
-0.026754304766654968,
-0.07044222950935364,
-0.03817464038729668,
0.06803727149963379,
-0.020253978669643402,
0.06852394342422485,
0.050640884786844254,
0.08850803226232529,
-0.04410488158464432,
0.02643391117453575,
0.04134922847151756,
0.03322779759764671,
-0.034088... |
5a834c4e-d4ba-5dfc-b752-5e607863d466 | /documentation/tutorials-lp-overview/ | Overview | documentation | false | Overview
### Basic Tutorials
*Get up and running with Qdrant in minutes.*
---
### Search Engineering Tutorials
*Master vector search modalities, reranking, and retrieval quality.*
---
### Operations & Scale
*Production-grade management, monitoring, and high-volume optimization.*
---
### Develop & Implement
*... | [
0.003171768970787525,
-0.08117932081222534,
-0.04613230377435684,
-0.012969465926289558,
0.062445152550935745,
-0.03955027461051941,
0.045476723462343216,
0.01073230430483818,
0.08478391170501709,
-0.01762661151587963,
0.03885901719331741,
-0.004196682013571262,
0.05178480222821236,
-0.041... | |
5a9a23ea-cf6a-506c-9e06-839b477e76d0 | /documentation/production-operations/access-control/ | Data Privacy with Qdrant's Role-Based Access Control (RBAC) | How Qdrant Handles Data Privacy and Security > Database API Keys > Payload Configuration | documentation | true | Data Privacy with Qdrant's Role-Based Access Control (RBAC) > How Qdrant Handles Data Privacy and Security > Database API Keys > Payload Configuration
There are several different options (claims) you can use in the JWT payload that help control access and functionality. Let’s look at them one by one; the [JWT configur... | [
0.033231135457754135,
-0.0155029296875,
-0.054461169987916946,
-0.04059210419654846,
0.06381383538246155,
-0.04019319266080856,
0.034064121544361115,
0.02018262818455696,
0.09329435229301453,
-0.038362275809049606,
0.011190208606421947,
0.008747726678848267,
0.07516905665397644,
-0.0143826... |
5a9c4bd7-a594-5f8a-9fc4-394ae1fc7715 | /course/essentials/day-7/camel/ | Integrating with Camel AI | CAMEL Auto-Retrieval Architecture | course | false | Integrating with Camel AI > CAMEL Auto-Retrieval Architecture
CAMEL (Communicative Agents for "Mind" Exploration of Large Language Model Society) provides an advanced framework for building multi-agent systems with automated RAG capabilities. The Auto-Retrieval module streamlines the process of expanding agent capabil... | [
0.032177139073610306,
-0.039949484169483185,
-0.048201899975538254,
-0.04432184621691704,
0.1090846136212349,
-0.05283195897936821,
0.07781063765287399,
0.04461987316608429,
0.004020040854811668,
-0.026830004528164864,
0.06849319487810135,
-0.016721835359930992,
0.0477420836687088,
-0.0573... |
5a9fac49-a06b-544c-b2c4-08a03a7cf96b | /articles/relevance-feedback/ | Relevance Feedback in Qdrant | Experiments > Parameters | article | false | Relevance Feedback in Qdrant > Experiments > Parameters
Training data size
| Dataset | Train queries | Test queries |
|------------|---------------:|-------------:|
| NFCorpus | 223 | 100 |
| FiQa-2018 | 500 | 148 |
| SCIDOCS | 500 | 500 |
| MSMARCO... | [
0.02916155941784382,
-0.04666450247168541,
-0.023891976103186607,
0.0018188527319580317,
0.046276263892650604,
-0.01975906267762184,
0.04828840121626854,
0.06897613406181335,
0.08099939674139023,
-0.018015995621681213,
-0.00851169042289257,
0.008084086701273918,
0.05654941499233246,
-0.062... |
5aa0940d-a3d1-5367-aef8-960be7ef57dd | /documentation/examples/rag-chatbot-vultr-dspy-ollama/ | Private RAG Information Extraction Engine | Deploying Qdrant Hybrid Cloud on Vultr > Qdrant Hybrid Cloud | documentation | true | Private RAG Information Extraction Engine > Deploying Qdrant Hybrid Cloud on Vultr > Qdrant Hybrid Cloud
Our [documentation](/documentation/hybrid-cloud/) contains a comprehensive guide on how to set up Qdrant in the Hybrid Cloud mode on Vultr. Please follow it carefully to get your Qdrant instance up and running. Onc... | [
0.04188084974884987,
-0.012438244186341763,
-0.057436708360910416,
-0.00931412074714899,
0.08433203399181366,
-0.04731195420026779,
0.07104387134313583,
0.06659310311079025,
0.0663239136338234,
-0.050407472997903824,
0.008949238806962967,
0.02854158543050289,
0.02558985911309719,
-0.062326... |
5aa0dc29-9ff4-5d37-b5ff-c6357a540342 | /documentation/tutorials-operations/time-based-sharding/ | Time-Based Sharding | Create Collection | documentation | true | Time-Based Sharding > Create Collection
Create a collection with [user-defined sharding](/documentation/scaling/distributed_deployment/#user-defined-sharding) by setting the sharding method to custom.
```python
from qdrant_client import models
collection_name = "my_collection"
if client.collection_exists(collection... | [
-0.012707255780696869,
-0.027915649116039276,
-0.059233810752630234,
-0.028792137280106544,
0.04436511918902397,
-0.06641809642314911,
0.023156847804784775,
0.06976083666086197,
0.07508082687854767,
-0.029231414198875427,
0.025566749274730682,
0.022107847034931183,
0.050127435475587845,
-0... |
5ab67222-35b0-5964-8728-71f2bc80a806 | /course/essentials/day-3/sparse-retrieval-demo/ | Demo: Keyword Search with Sparse Vectors | Sparse Neural Retrieval > Beyond SPLADE > Qdrant's Sparse Neural Retrievers | course | false | Demo: Keyword Search with Sparse Vectors > Sparse Neural Retrieval > Beyond SPLADE > Qdrant's Sparse Neural Retrievers
- **When to use miniCOIL:** you need **exact keyword matches** in the retrieved results, as you're working in a domain where precise term overlap matters (e.g., e-commerce, legal, medical), but want r... | [
0.04774583876132965,
-0.05715750902891159,
-0.05515717715024948,
-0.037229273468256,
0.06995975971221924,
-0.04875500500202179,
0.08689074218273163,
0.029385045170783997,
0.06997349113225937,
-0.024391330778598785,
0.019723419100046158,
0.026440424844622612,
0.06975020468235016,
0.00238404... |
5ab96d21-a52a-5fc1-959f-bbe5b577f862 | /articles/scalar-quantization/ | Scalar Quantization for Vector Search | Benchmarks > Gist-960 | article | false | Scalar Quantization for Vector Search > Benchmarks > Gist-960
| | Upload and indexing time | Mean search precision, ef = 128 | Mean search time, ef = 128 | Mean search precision, ef = 256 | Mean search time, ef = 256 | Mean search precision, ef = 512 | Mean search time, ef = 512 |
|---|---|---|---|---|---|---|---|
| N... | [
0.014475494623184204,
-0.030604897066950798,
-0.030143894255161285,
-0.018224604427814484,
0.03092244267463684,
-0.03374665230512619,
0.054255060851573944,
0.06438126415014267,
0.12684063613414764,
-0.05875197425484657,
0.03181397169828415,
0.02841222658753395,
0.04597926139831543,
-0.0650... |
5adec9f8-a668-5ce6-b617-825b8a5e6e0d | /documentation/tutorials-search-engineering/branch-aware-search/ | Branch-Aware Search | Scaling to Production | documentation | false | Branch-Aware Search > Scaling to Production
- **The filter grows with lineage depth, not corpus size.** Each query walks the ancestry once, adding one candidate clause and one exclusion clause per ancestor. The filter's size tracks how deep a branch sits, not how many files or versions the collection holds.
- **Qdran... | [
0.035525593906641006,
-0.02992837131023407,
-0.06975018978118896,
-0.031216200441122055,
0.058300357311964035,
-0.02113012783229351,
0.05277719721198082,
0.03502020984888077,
0.0838828980922699,
-0.04083719477057457,
0.04283102974295616,
0.02527342364192009,
0.06739000231027603,
-0.0228669... |
5b0327d2-8a12-54b8-93e2-143ca72f0b3d | /course/essentials/day-2/pitstop-project/ | Project: HNSW Performance Benchmarking | course | false | Project: HNSW Performance Benchmarking
Day 2
Now that you've seen how [HNSW](/articles/filterable-hnsw/) parameters and payload indexes affect performance with the DBpedia dataset, it's time to optimize for your own domain and use case. | [
0.02861294150352478,
-0.025979341939091682,
-0.056505247950553894,
-0.05951414257287979,
0.040447212755680084,
-0.06896766275167465,
0.07148697227239609,
0.0659928023815155,
0.054254479706287384,
-0.017060089856386185,
0.020671334117650986,
0.04273238405585289,
0.03407096490263939,
-0.0330... | |
5b0bbfae-938b-5dad-8a5d-85a63cb1f1e9 | /documentation/tutorials-develop/search-as-you-type/ | Semantic Search As You Type | Performance | documentation | false | Semantic Search As You Type > Performance
The original service was benchmarked in 2023 on a local AMD Ryzen 9 5900HX with 16GB of RAM, comparing the previous Python service (FastAPI, SentenceTransformers) with the Rust service (Actix Web, ONNX Runtime). The table shows the average time and error bound, for up to a tho... | [
0.03438088670372963,
-0.056381020694971085,
-0.04939344897866249,
0.0037024598568677902,
0.05181438848376274,
-0.023248761892318726,
0.03262859210371971,
-0.0020309321116656065,
0.05862343683838844,
-0.03231145069003105,
0.07962621003389359,
-0.003339665476232767,
0.03673991560935974,
-0.0... |
5b16fb11-93f7-5590-98b8-e330f83730c1 | /documentation/examples/qdrant-dspy-medicalbot/ | Building a Chain-of-Thought Medical Chatbot with Qdrant and DSPy | documentation | false | Building a Chain-of-Thought Medical Chatbot with Qdrant and DSPy
Accessing medical information from LLMs can lead to hallucinations or outdated information. Relying on this type of information can result in serious medical consequences. Building a trustworthy and context-aware medical chatbot can solve this.
In this ... | [
0.037274811416864395,
-0.044875286519527435,
-0.06668218970298767,
-0.03994635492563248,
0.07308792322874069,
-0.041687171906232834,
0.04772520437836647,
0.046849969774484634,
0.042942147701978683,
-0.03414379060268402,
0.034752074629068375,
0.015634572133421898,
0.06284891068935394,
-0.05... | |
5b189693-7133-5d3e-b5cc-cb1173c6dabe | /documentation/ops-optimization/optimizer/ | Optimizer | Optimization Monitoring | documentation | true | Optimizer > Optimization Monitoring
*Available as of v1.17.0*
The `/collections/{collection_name}/optimizations` API endpoint returns information about the optimization of a specific collection, including:
- A summary of optimization activity, with the number of queued optimizations, queued segments, queued points, a... | [
-0.001452839933335781,
-0.03821392357349396,
-0.01960008218884468,
-0.04484320804476738,
0.0469721294939518,
-0.027844488620758057,
0.0248822383582592,
-0.0010285155149176717,
0.08463980257511139,
-0.0565851554274559,
0.055657822638750076,
0.017677387222647667,
0.04584208130836487,
-0.0587... |
5b1be092-927e-5f9c-a6e0-d64b94f4ca14 | /documentation/platforms/n8n/ | N8N | documentation | false | N8N
[N8N](https://n8n.io/) is an automation platform that allows you to build flexible workflows focused on deep data integration.
[Qdrant's official node](https://github.com/qdrant/n8n-nodes-qdrant) for n8n enables semantic search capabilities in your workflows. | [
0.015755711123347282,
-0.04959367960691452,
-0.06771145761013031,
-0.023117758333683014,
0.047795914113521576,
-0.04888613894581795,
0.05621420592069626,
0.04689946398139,
0.0707850381731987,
-0.011466856114566326,
0.08768307417631149,
0.01327064074575901,
0.02237454243004322,
-0.048605315... | |
5b35b24b-7a6a-5d0b-9122-66695ec8d414 | /articles/what-are-embeddings/ | Vector Embeddings Explained: How They Work in ML & Search | How do embeddings work? > Creating vector embeddings | article | false | Vector Embeddings Explained: How They Work in ML & Search > How do embeddings work? > Creating vector embeddings
Embeddings translate the complexities of human language to a format that computers can understand. It uses neural networks to assign **numerical values** to the input data, in a way that similar data has si... | [
0.06550094485282898,
-0.0345279686152935,
-0.08293133974075317,
-0.022098898887634277,
0.05512566119432449,
-0.018038291484117508,
0.048356570303440094,
0.03082725591957569,
0.06663907319307327,
-0.023137904703617096,
0.04708520323038101,
0.015535044483840466,
0.07583221048116684,
-0.03715... |
5b39d496-1543-5249-8291-261afceaccdb | /course/multi-vector-search/module-1/late-interaction-basics/ | Late Interaction Basics | Late Interaction: The Core Paradigm > How It Works | course | false | Late Interaction Basics > Late Interaction: The Core Paradigm > How It Works
Instead of compressing a document into a single vector, late interaction represents it as a collection of contextualized token embeddings
1. **Encode:** Pass the document through an encoder (like ColBERT) to generate one embedding vector per... | [
0.03854943439364433,
-0.04838244244456291,
-0.057068996131420135,
-0.05443535000085831,
0.055414311587810516,
-0.017256250604987144,
0.025063905864953995,
0.025409365072846413,
0.08395180106163025,
-0.014884759671986103,
0.02725515514612198,
-0.02204977534711361,
0.03196738287806511,
-0.00... |
5b5d4ece-af8a-5475-b432-dc45efbc6463 | /documentation/snapshots/ | Snapshots | Storage > Local file system | documentation | true | Snapshots > Storage > Local file system
By default, snapshots are stored at `./snapshots` or at `/qdrant/snapshots` when using our Docker image.
The target directory can be controlled through the [configuration](/documentation/ops-configuration/configuration/):
```yaml
storage:
# Specify where you want to store sn... | [
0.029227467253804207,
-0.030067389830946922,
-0.020305782556533813,
-0.03536259010434151,
0.01254392508417368,
-0.02986028976738453,
0.04102478176355362,
0.02276839315891266,
0.07788495719432831,
-0.01224436517804861,
0.0156402587890625,
-0.026465171948075294,
0.06644566357135773,
-0.01490... |
5b60b534-8dd0-5fe8-8ef1-0260144eee82 | /documentation/tutorials-operations/large-scale-search/ | Large-Scale Search | Search | documentation | true | Large-Scale Search > Search
After the dataset is uploaded and indexed, we can start searching for similar vectors.
We can start by exploring the dataset in Web-UI. So you can get an intuition into the search performance, not just table numbers.
Web-UI default requests do not use oversampling, but the observable resu... | [
0.040894895792007446,
-0.07807740569114685,
-0.0215649101883173,
0.0000480564558529295,
0.07465248554944992,
-0.004335508681833744,
0.08059587329626083,
0.01618216559290886,
0.06966152042150497,
-0.04871075600385666,
0.023165656253695488,
0.02639716863632202,
0.058579664677381516,
-0.03930... |
5b6f11f2-a158-5e28-91b1-cb82e145787d | /articles/dataset-quality/ | Detecting Dataset Errors with Similarity Search | Conclusion | article | false | Detecting Dataset Errors with Similarity Search > Conclusion
In this article, we enlightened distance-based methods to find errors in categorized datasets.
Showed how to find incorrectly placed items in the furniture web store.
I hope these methods will help you catch sneaky samples leaked into the wrong categories in... | [
0.09054487198591232,
0.007197552360594273,
-0.03570373356342316,
-0.049406155943870544,
0.04799432307481766,
-0.0235399529337883,
0.06166333332657814,
0.034072037786245346,
0.05609873682260513,
0.006688477471470833,
0.02196558564901352,
0.0702945813536644,
0.06004464626312256,
0.0055780331... |
5b7ad51a-eeea-59ce-84c2-746ce62178a3 | /documentation/security/ | Security | Authentication > Authenticate with an API Key | documentation | true | Security > Authentication > Authenticate with an API Key
To authenticate with an API key, whether it's an admin key, read-only key, or granular access token, provide it in the `api-key` request header:
```python
from qdrant_client import QdrantClient
client = QdrantClient(
url="https://xyz-example.eu-central.aws... | [
-0.009828790090978146,
-0.025642860680818558,
-0.047266650944948196,
-0.07761944085359573,
0.05728210136294365,
-0.06855439394712448,
0.009516348131000996,
0.08105044066905975,
0.04063468798995018,
-0.007251190021634102,
0.024079149588942528,
0.05142438784241676,
0.07157599925994873,
-0.02... |
5b7daa85-4f8b-5717-8017-4802a396933c | /documentation/hybrid-cloud/hybrid-cloud-setup/ | Setup Hybrid Cloud | documentation | false | Setup Hybrid Cloud
The following instruction set will show you how to properly set up a **Qdrant cluster** in your **Hybrid Cloud Environment**.
To learn how Hybrid Cloud works, [read the overview document](/documentation/hybrid-cloud/). | [
0.02540396712720394,
-0.004970427602529526,
-0.04654383659362793,
-0.050179190933704376,
0.0776183009147644,
-0.02422523684799671,
0.036464300006628036,
0.024139389395713806,
0.0497872494161129,
-0.005979446228593588,
0.008906149305403233,
-0.0010723761515691876,
0.04829299449920654,
-0.03... | |
5b7e1dfa-863f-50d6-85d6-ec0b7803c4bc | /documentation/production-operations/access-control/ | Data Privacy with Qdrant's Role-Based Access Control (RBAC) | Conclusion | documentation | false | Data Privacy with Qdrant's Role-Based Access Control (RBAC) > Conclusion
Vector similarity is increasingly becoming the backbone of AI applications that leverage unstructured data. By transforming data into vectors – their numerical representations – organizations can build powerful applications that harness semantic ... | [
0.041283685714006424,
-0.0533217154443264,
-0.06709761172533035,
-0.020944960415363312,
0.05324608087539673,
-0.047832924872636795,
0.09014622867107391,
0.05157453566789627,
0.05990532785654068,
-0.031367070972919464,
0.029888220131397247,
0.040474798530340195,
0.06775517016649246,
-0.0179... |
5b903e58-6957-5865-ad8f-b76195bff6f1 | /course/beginners/module-1/why-similarity-is-not-enough/ | Why Similarity Alone Is Not Enough | Exact Matching: When You Need a Filter, Not Similarity | course | true | Why Similarity Alone Is Not Enough > Exact Matching: When You Need a Filter, Not Similarity
Not every query needs semantic understanding. A query for an exact product code, like "SKU-48291", needs an exact match instead. Try it yourself: embed the query and three candidate product codes, then compare their cosine scor... | [
0.04441526532173157,
-0.038526151329278946,
-0.08490043878555298,
-0.04774957150220871,
0.07257729023694992,
-0.028471412137150764,
0.031237462535500526,
0.024287089705467224,
0.08643271774053574,
-0.03904972970485687,
0.04815879836678505,
0.024039311334490776,
0.07788722217082977,
-0.0289... |
5b93fadd-06d8-57ac-8ef9-62218ef492c9 | /documentation/agentic-tools/prompts/ | Prompts in the Docs | documentation | false | Prompts in the Docs
Qdrant documentation includes short, ready-to-use prompts on the pages where they are relevant: setting up monitoring on the monitoring page, diagnosing memory issues on the optimization page, and so on.
Each prompt is collapsed by default, requires no additional input, and tells your assistant wh... | [
0.004992519039660692,
-0.057280100882053375,
-0.03773718327283859,
-0.03955931216478348,
0.07312114536762238,
-0.05137680470943451,
0.04551905021071434,
0.048933032900094986,
0.09462017565965652,
-0.02476218342781067,
0.03993277996778488,
-0.01526590809226036,
0.056955840438604355,
-0.0406... | |
5b9e6588-0456-5de7-a8a4-033ec86e7a10 | /documentation/tutorials-basics/multimodal-search/ | Multimodal Search | Overview | documentation | false | Multimodal Search > Overview
You often understand and share information more effectively when combining different types of data. The taste of comfort food can trigger childhood memories. A song might be described with just "pam pam clap" sounds instead of a paragraph. Emojis and stickers can express a feeling or a com... | [
0.06345944851636887,
-0.03138595446944237,
-0.04823281615972519,
-0.03600475937128067,
0.07088740170001984,
-0.028526240959763527,
0.08267571032047272,
0.016154441982507706,
0.03692365065217018,
-0.01769435778260231,
0.03373135253787041,
-0.024192292243242264,
0.059167321771383286,
-0.0517... |
5ba0803b-941a-5275-b956-40cc54938b7c | /documentation/edge/edge-api/snapshots/ | Snapshots | documentation | false | Snapshots
Snapshots move data between an Edge Shard and a Qdrant server collection. This section covers the methods; for how to combine them, refer to [Data Synchronization Patterns](/documentation/edge/edge-data-synchronization-patterns/). | [
0.007054266519844532,
-0.022243427112698555,
-0.07348017394542694,
-0.06285320967435837,
0.02320871502161026,
-0.017896294593811035,
0.11878085881471634,
0.01690354011952877,
0.10405029356479645,
-0.009070539847016335,
0.045966748148202896,
0.01864108070731163,
0.0486433319747448,
0.020151... | |
5bbcb583-6658-5d1f-b602-33b0f4c7e96f | /articles/predicting-weak-retrieval/ | Predicting Weak Retrieval Without an LLM | Turn the Signal into a Gate | article | true | Predicting Weak Retrieval Without an LLM > Turn the Signal into a Gate
A threshold turns the signal into a decision. Orient the chosen signal so lower means weaker, negating the ones that fire high on weak retrieval (like `retriever_divergence`), so a single floor decides: below it, treat the retrieval as weak and esc... | [
0.054614488035440445,
-0.04966655373573303,
-0.04602208361029625,
-0.038599640130996704,
0.04101106524467468,
-0.0023607255425304174,
0.01460824441164732,
0.010515675880014896,
0.048187900334596634,
-0.004925248678773642,
-0.02243630774319172,
-0.00006710515299346298,
0.03558316454291344,
... |
5bc9fcdd-0ee8-5ce8-aca7-7b79d4980e47 | /documentation/inference/cloud-inference/ | Cloud Inference | Automatic Query and Passage Prefixes | documentation | false | Cloud Inference > Automatic Query and Passage Prefixes
Some embedding models are trained to expect different text prefixes for query inputs versus passage or document inputs. Using the wrong prefix or omitting one can degrade retrieval quality.
Qdrant Cloud Inference automatically applies the correct prefix for the m... | [
0.0666174367070198,
-0.04338514432311058,
-0.007710229605436325,
-0.04404061287641525,
0.07668942958116531,
-0.04042406007647514,
0.06162029877305031,
0.03937949240207672,
0.029076768085360527,
-0.02470666542649269,
0.010402283631265163,
0.005959516391158104,
0.02973216027021408,
-0.023296... |
5bcab969-fb8d-5b45-a3d3-a7ad4e3bbc5c | /documentation/embeddings/jina-embeddings/ | Jina AI | Embed Documents | documentation | true | Jina AI > Embed Documents
One request embeds a list of inputs. Set `task` to `retrieval.passage` for documents. Each item in the response carries the `index` of its input, which this example uses as the point ID:
```python
documents = [
"Qdrant is a vector search engine.",
"Jina AI embedding models turn text,... | [
0.019217709079384804,
-0.05872649699449539,
-0.0336260050535202,
-0.048910535871982574,
0.07959667593240738,
-0.03079252503812313,
0.06280442327260971,
0.06339500099420547,
0.06619596481323242,
-0.013331635855138302,
0.020230546593666077,
0.030479054898023605,
0.030338821932673454,
-0.0278... |
5bf58140-25a8-5682-af61-2ba5a5c478ed | /documentation/tutorials-basics/generate-embeddings-fastembed/ | FastEmbed: Qdrant's Efficient Python Library for Embedding Generation | documentation | true | FastEmbed: Qdrant's Efficient Python Library for Embedding Generation
| Time: 10 min | Level: Beginner |
| --- | ----------- |
[FastEmbed](https://github.com/qdrant/fastembed) is a Python library that generates embeddings on CPU, with no GPU, PyTorch, or API key required. In this tutorial, you generate embeddings for... | [
0.023740049451589584,
-0.019681518897414207,
-0.023044170811772346,
-0.03729498013854027,
0.0860021635890007,
-0.053849563002586365,
0.08729621767997742,
0.05139005929231644,
0.0575491301715374,
-0.04326104000210762,
0.027977701276540756,
0.018156692385673523,
0.04489123076200485,
-0.05202... | |
5c3538bc-aec0-56d4-ad3f-69fff7333da5 | /documentation/migrate-to-qdrant/from-qdrant/ | From Qdrant | What You Need | documentation | false | From Qdrant > What You Need
- **Source Qdrant URL** — gRPC endpoint of the source instance
- **Target Qdrant URL** — gRPC endpoint of the target instance
- **Source collection name**
- **Target collection name** (must be different from source if using the same instance)
- **API keys** — for each instance, if authentic... | [
0.009658346883952618,
-0.04341380298137665,
-0.04422939568758011,
-0.010441766120493412,
0.06624159216880798,
-0.025058500468730927,
0.06197729334235191,
0.02594277635216713,
0.07599609345197678,
-0.04563302919268608,
0.022444812580943108,
-0.006720856297761202,
0.04388716071844101,
-0.043... |
5c4ee4a0-0f19-53ea-918b-5c9aa43c777a | /course/beginners/module-1/module-overview/ | Module Overview | course | true | Module Overview
Module 1
Traditional search retrieves documents by matching the words in a query to the words in an index. It's fast and still useful for precise terms, but it can miss relevant results when people express the same intent differently. In this module, you'll see how embeddings let search compare meanin... | [
0.0399320162832737,
-0.04976896569132805,
-0.0687345415353775,
-0.013528755865991116,
0.07630689442157745,
-0.008303383365273476,
0.05107397213578224,
0.026012441143393517,
0.08500763773918152,
-0.01709277369081974,
0.03244553506374359,
0.020773377269506454,
0.04510045424103737,
-0.0081916... | |
5c5a5ff2-f332-553e-ad0d-d3a4e721dcf2 | /articles/relevance-feedback/ | Relevance Feedback in Qdrant | Relevance Feedback Query > How to Use It | article | false | Relevance Feedback in Qdrant > Relevance Feedback Query > How to Use It
For ease of use, as one shouldn't need a machine learning degree to use a new feature, we published a [Python package that customizes Naive Formula weights for your dataset, retriever, and feedback model](https://pypi.org/project/qdrant-relevance-... | [
0.06820587068796158,
-0.025804070755839348,
-0.03909366950392723,
-0.07748513668775558,
0.06160309538245201,
-0.03898340091109276,
0.046585503965616226,
0.02751847170293331,
0.05156616494059563,
-0.022796018049120903,
0.015336194075644016,
0.018175844103097916,
0.0873507708311081,
-0.02282... |
5c5d4d0e-db03-5fb1-98f0-0df2ab9956cd | /documentation/manage-data/collections/ | Collections | Create a Collection | documentation | true | Collections > Create a Collection
```python
from qdrant_client import QdrantClient, models
client = QdrantClient(url="http://localhost:6333")
client.create_collection(
collection_name="{collection_name}",
vectors_config=models.VectorParams(size=100, distance=models.Distance.COSINE),
)
```
In addition to the... | [
0.015767039731144905,
-0.0491240955889225,
-0.051191650331020355,
-0.027230409905314445,
0.06532184779644012,
-0.05270032957196236,
0.020945416763424873,
0.058341704308986664,
0.08698208630084991,
-0.032016173005104065,
0.01623217947781086,
0.009239330887794495,
0.046941813081502914,
-0.04... |
5c7b1d6c-787b-5d25-a147-b4837cef5480 | /course/essentials/day-2/collection-tuning-demo/ | Demo: HNSW Performance Tuning | Step 6: Enable HNSW Indexing | course | true | Demo: HNSW Performance Tuning > Step 6: Enable HNSW Indexing
Now switch from `m=0` to `m=16` to build HNSW connections and improve search time.
```python
client.update_collection(
collection_name=collection_name,
hnsw_config=models.HnswConfigDiff(
m=16 # Build HNSW now: m=16 after the bulk load.
... | [
0.03615293651819229,
-0.05121971294283867,
-0.07401730865240097,
0.015089420601725578,
0.04612494632601738,
-0.05518891289830208,
0.04452187567949295,
0.040624670684337616,
0.05482058227062225,
0.007594672963023186,
0.03120439313352108,
0.02177397720515728,
0.06998150795698166,
-0.01295464... |
5cb5cd09-313c-5569-b7bf-52dab5f4b00c | /documentation/hybrid-cloud/hybrid-cloud-setup/ | Setup Hybrid Cloud | Prerequisites | documentation | false | Setup Hybrid Cloud > Prerequisites
Qdrant Hybrid Cloud is available as part of our Enterprise plan. To get access to Qdrant Hybrid Cloud, please [contact us](/contact-us/).
- **Kubernetes Cluster:** To create a Hybrid Cloud Environment, you need a [standard compliant](https://www.cncf.io/training/certification/softwa... | [
0.011705737560987473,
-0.034327466040849686,
-0.051971811801195145,
-0.037752605974674225,
0.06918715685606003,
-0.050626590847969055,
0.07158487290143967,
0.04608730971813202,
0.057317014783620834,
-0.056824881583452225,
0.022615492343902588,
-0.010989544913172722,
0.023007720708847046,
-... |
5cba161c-2c3f-54ce-9490-114718491a89 | /documentation/search-precision/reranking-semantic-search/ | Reranking for Better Search | Implementing Vector Search with Reranking > Implementation > Setup | documentation | true | Reranking for Better Search > Implementing Vector Search with Reranking > Implementation > Setup
To follow along with this tutorial, you'll need a few key tools:
- Python Client for Qdrant
- Cohere
Let’s install everything you need in one go using the Python package manager:
```shell
pip install qdrant-client coher... | [
0.04283273592591286,
-0.07921089977025986,
-0.061156999319791794,
-0.026927301660180092,
0.06070735305547714,
-0.03947307914495468,
0.09463182091712952,
0.021788574755191803,
0.07486941665410995,
-0.0331803523004055,
0.013641548343002796,
0.0268867164850235,
0.05300668627023697,
-0.0282100... |
5ccb0519-0172-555f-9793-7206180fdd07 | /documentation/manage-data/indexing/ | Indexing | Full-Text Index > Tokenizers > Code | documentation | false | Indexing > Full-Text Index > Tokenizers > Code
*Available as of v1.20.0*
The `code` tokenizer is designed for source code. It splits identifiers into their parts, so a name matches regardless of its naming convention. For example, filtering on `getUserName`, `get_user_name`, or `user name` finds text that contains an... | [
0.02690032683312893,
-0.03207187354564667,
-0.07077675312757492,
-0.02544313296675682,
0.04401129111647606,
-0.012500434182584286,
0.028391456231474876,
0.019898053258657455,
0.038021624088287354,
0.015755625441670418,
0.045610833913087845,
0.008050684817135334,
0.04649097099900246,
-0.038... |
5ccd7ee9-ebba-552d-8844-f493a4ca18a3 | /documentation/manage-data/points/ | Points | Scroll Points | documentation | true | Points > Scroll Points
Sometimes it might be necessary to get all stored points without knowing IDs, or iterate over points that correspond to a filter.
REST API ([Schema](https://api.qdrant.tech/master/api-reference/points/scroll-points)):
```python
client.scroll(
collection_name="{collection_name}",
scroll... | [
0.050671275705099106,
0.008780119009315968,
-0.018461301922798157,
-0.024935461580753326,
0.047800175845623016,
-0.05388546362519264,
0.04704498499631882,
0.03289419785141945,
0.06167401745915413,
-0.015860222280025482,
0.04308789223432541,
0.05846896395087242,
0.01222133543342352,
-0.0233... |
5cddaa68-614a-5b72-a231-0c088ed9a424 | /course/beginners/module-5/signal-sources-and-embedding-models/ | Signal Sources and Embedding Models | course | true | Signal Sources and Embedding Models
Module 5
Two models cover every signal here, and both run through FastEmbed exactly as in Modules 3 and 4: name the model, pass the content, and the client embeds it locally before upload.
| Signal source | Modality | Embedding model | Vectors it produces |
|---------------|------... | [
0.06452468782663345,
-0.02129100076854229,
-0.04258883744478226,
-0.05508282408118248,
0.07576465606689453,
-0.01841953583061695,
0.08101051300764084,
0.0700594037771225,
0.019900891929864883,
-0.015983857214450836,
0.02030990831553936,
0.012131846509873867,
0.027407562360167503,
-0.055744... | |
5d0530bf-fa2b-519d-a10e-5959837bcb2d | /documentation/faq/qdrant-fundamentals/ | Qdrant Fundamentals | Search > How can I search without a vector? | documentation | false | Qdrant Fundamentals > Search > How can I search without a vector?
You are likely looking for the [scroll](/documentation/manage-data/points/#scroll-points) method. It allows you to retrieve the records based on filters or even iterate over all the records in the collection. | [
0.03133983537554741,
-0.03769899532198906,
-0.05467510595917702,
-0.014498153701424599,
0.05815356969833374,
-0.01605169288814068,
0.02825760282576084,
0.02732321433722973,
0.06718478351831436,
0.019440477713942528,
0.02175660990178585,
0.08271145820617676,
0.08036065846681595,
-0.01050327... |
5d07b405-8147-56c4-8faa-fadfb950133e | /documentation/manage-data/multitenancy/ | Multitenancy | Partition by Payload > Per-Tenant Scoring Statistics | documentation | false | Multitenancy > Partition by Payload > Per-Tenant Scoring Statistics
- `idf` defaults to `global` (shard-wide statistics), the same as omitting it.
- [Create a payload index](/documentation/manage-data/indexing/#create-a-payload-index) and/or [tenant index](/documentation/manage-data/indexing/#tenant-index) for any fie... | [
0.039612844586372375,
-0.0025388195645064116,
-0.05782133340835571,
-0.027890855446457863,
0.04391263425350189,
-0.03238428384065628,
0.05992162227630615,
0.032879553735256195,
0.09188858419656754,
-0.004824324045330286,
0.03229154646396637,
0.01925533078610897,
0.031125009059906006,
-0.01... |
5d100c48-c6c4-58d7-98d0-dd303de79abc | /documentation/manage-data/quantization/ | Quantization | Setting Up Quantization in Qdrant | documentation | false | Quantization > Setting Up Quantization in Qdrant
You can configure quantization for a collection by specifying the quantization parameters in the `quantization_config` section of the collection configuration.
Quantization will be automatically applied to all vectors during the indexation process.
Quantized vectors ar... | [
0.025262009352445602,
-0.06472635269165039,
-0.01913459040224552,
-0.042676251381635666,
0.01933581940829754,
-0.025706252083182335,
0.02799118496477604,
0.05350697785615921,
0.08773917704820633,
-0.032412610948085785,
0.021137025207281113,
0.014855673536658287,
0.07347345352172852,
-0.043... |
5d10ca38-9b2f-5d5f-af88-227a22d467e3 | /course/essentials/day-7/tensorlake/ | Integrating with Tensorlake | TensorLake Knowledge Graph Integration > Qdrant Integration Workflow | course | false | Integrating with Tensorlake > TensorLake Knowledge Graph Integration > Qdrant Integration Workflow
The complete integration process follows these steps:
1. **Document Processing with TensorLake**:
- Parse documents to extract structured information
- Generate knowledge graphs with preserved relationships
- C... | [
0.014365519396960735,
-0.04047570750117302,
-0.07475239783525467,
-0.008498130366206169,
0.08838611096143723,
-0.0328977070748806,
0.039914820343256,
0.042799945920705795,
0.04840646684169769,
-0.027821218594908714,
0.01498884242027998,
0.02066897600889206,
0.056887898594141006,
-0.0145757... |
5d2b79c6-962c-501f-9fba-f5b012addb38 | /course/multi-vector-search/module-3/evaluating-pipelines/ | Evaluating Search Pipelines | course | false | Evaluating Search Pipelines
Module 3
Throughout this module, you've learned many optimization techniques: quantization to reduce memory, pooling to compress representations, MUVERA for efficient indexing, and multi-stage retrieval to balance speed with accuracy. But how do you know which combination is right for *you... | [
0.04071933776140213,
-0.02336011454463005,
-0.049889203161001205,
-0.05831563100218773,
0.0732787474989891,
-0.026093455031514168,
0.05234043672680855,
-0.0013596309581771493,
0.08107669651508331,
-0.026738306507468224,
0.010445392690598965,
0.022328989580273628,
0.05050018057227135,
-0.03... | |
5d2fbc50-f5df-597e-89b6-84c6241f0cfe | /documentation/tutorials-search-engineering/multi-representation-search/ | Multi-Representation Search | documentation | false | Multi-Representation Search
| Time: 45 min | Level: Intermediate | Output: [GitHub](https://github.com/qdrant/examples/blob/master/multi-representation-search/multi-representation-search.ipynb) | [](https://githubtocolab.com/qdrant/examples/blob... | [
0.018439551815390587,
-0.03508872166275978,
-0.05608811601996422,
-0.04248715937137604,
0.08740394562482834,
-0.05676911771297455,
0.047144901007413864,
0.05646368861198425,
0.06245145946741104,
-0.035980116575956345,
0.03496136516332626,
0.014665758237242699,
0.0669199675321579,
-0.049124... | |
5d435afa-5b7e-52a1-bd14-7629077ea819 | /documentation/security/ | Security | Authentication > Admin API Key > Enforce Authentication on Internal Communication | documentation | true | Security > Authentication > Admin API Key > Enforce Authentication on Internal Communication
*Available as of v1.18.0. Enabled by default as of v1.20.0.*
In a distributed deployment, peers talk to each other over the internal gRPC API (port 6335 by default). Every peer attaches its admin `api_key` to outgoing interna... | [
0.0461348220705986,
-0.02871357835829258,
-0.06830480694770813,
-0.017220497131347656,
0.043093785643577576,
-0.02580632083117962,
0.03065519966185093,
0.07078145444393158,
0.05298517644405365,
-0.010619567707180977,
0.023605626076459885,
0.02824384905397892,
0.05638770014047623,
-0.068820... |
5d51b809-e38e-5a51-a0be-7075820ac3e4 | /documentation/tutorials-build-essentials/question-answering-langchain/ | Question Answering with LangChain and Qdrant | How to Implement Question Answering with LangChain and Qdrant > Step 4: Testing out the chain | documentation | true | Question Answering with LangChain and Qdrant > How to Implement Question Answering with LangChain and Qdrant > Step 4: Testing out the chain
And that's it! We can put in some queries, and LangChain will perform all the required processing to find the answer in the provided context. Since we
already have a `questions` ... | [
0.03430113196372986,
-0.05448610708117485,
-0.03185877576470375,
-0.016839640215039253,
0.055049192160367966,
-0.05552559345960617,
0.0533340759575367,
0.021389933302998543,
0.05766402557492256,
-0.05035649240016937,
0.022150585427880287,
0.011097400449216366,
0.0567445233464241,
-0.040316... |
5d5ca68b-f4b8-5860-9536-92ac43bd88fc | /documentation/data-synchronization/with-postgres/ | Keeping Postgres and Qdrant in Sync | Tier 3: Change Data Capture with Debezium + Redpanda > The Killer Feature: Replaying History | documentation | false | Keeping Postgres and Qdrant in Sync > Tier 3: Change Data Capture with Debezium + Redpanda > The Killer Feature: Replaying History
CDC's most underappreciated advantage is replayability. If Qdrant needs to be rebuilt from scratch — new index configuration, migration to a new cluster, disaster recovery — you can do it ... | [
0.0370575487613678,
-0.05057791993021965,
-0.029454505071043968,
-0.06312283128499985,
0.05562564358115196,
-0.051021676510572433,
0.07164926081895828,
0.04174507409334183,
0.059751737862825394,
0.0017002958338707685,
0.032058507204055786,
0.012601287104189396,
0.041623204946517944,
-0.038... |
5d84f70c-4c97-5ea4-9e3c-d2ac8f2e6039 | /documentation/manage-data/vectors/ | Vectors | Vector Storage | documentation | false | Vectors > Vector Storage
Qdrant always [stores vectors on disk](/documentation/manage-data/storage/#vector-storage). To improve query latency, you can also load a copy of the vectors into memory for faster access by configuring [a memory tier](/documentation/ops-configuration/memory-tiers/). | [
0.030717262998223305,
-0.038279835134744644,
-0.012407869100570679,
-0.04523928835988045,
0.03040587715804577,
-0.037659671157598495,
0.06246151775121689,
0.012644038535654545,
0.09936244785785675,
-0.029355322942137718,
0.002167291007936001,
0.009816795587539673,
0.08846067637205124,
0.01... |
5d8e80b9-edd4-5128-be1f-991b271b4d86 | /documentation/data-management/confluent/ | Confluent Kafka | Message Formats | documentation | true | Confluent Kafka > Message Formats
The connector supports messages in the following formats.
_Click each to expand._
Unnamed/Default vector
Reference: [Creating a collection with a default vector](/documentation/manage-data/collections/#create-a-collection).
```json
{
"collection_name": "{collection_name}",
... | [
0.008461854420602322,
-0.05013393238186836,
-0.040322303771972656,
-0.04902014136314392,
0.04434024542570114,
-0.036341793835163116,
0.0689593106508255,
0.09483542293310165,
0.04424641653895378,
-0.04329856485128403,
-0.009771548211574554,
0.010950274765491486,
0.024767013266682625,
-0.024... |
5d8ead7f-a912-5d6b-8d0a-2968c93f0ef9 | /documentation/production-operations/vector-search-in-production/ | Vector Search in Production: Scaling, HA & Tuning Guide | Conclusion | documentation | false | Vector Search in Production: Scaling, HA & Tuning Guide > Conclusion
In conclusion, **vector search in production** isn't tied to a specific cloud provider or infrastructure. The same core principles of **careful configuration, robust ingestion/indexing, intelligent scaling, thorough backups, strong observability, and... | [
-0.03182916343212128,
-0.04902008920907974,
-0.061137713491916656,
-0.029645336791872978,
0.05247192084789276,
0.003171303076669574,
0.07703844457864761,
0.030438309535384178,
0.08105562627315521,
-0.006485603284090757,
0.010953865014016628,
0.02658000774681568,
0.050322312861680984,
-0.02... |
5da2a349-6c2d-5c45-a93e-563e16e6354b | /articles/search-feedback-loop/ | Relevance Feedback in Information Retrieval | So, what are the takeaways? | article | false | Relevance Feedback in Information Retrieval > So, what are the takeaways?
Pseudo Relevance Feedback (PRF) is known to improve the effectiveness of lexical retrievers. Several PRF-based approaches – mainly query terms expansion-based – are successfully integrated into traditional retrieval systems. At the same time, th... | [
0.051220692694187164,
-0.01122216321527958,
-0.08722523599863052,
-0.06385330855846405,
0.05212734639644623,
-0.03460503742098808,
0.011828766204416752,
0.03345458209514618,
0.05774775519967079,
-0.015392075292766094,
0.007149612996727228,
0.00986336451023817,
0.0781761109828949,
-0.036032... |
5ddeb543-7acc-51bc-93ac-2619d22b9754 | /course/multi-vector-search/module-3/multi-stage-retrieval/ | Multi-Stage Retrieval with Universal Query API | What's Next | course | false | Multi-Stage Retrieval with Universal Query API > What's Next
You've learned how to build sophisticated multi-stage retrieval pipelines that combine different techniques for optimal results. But multi-vector representations can be memory-intensive, especially at scale.
In the next lesson, you'll discover **vector quan... | [
0.03634665533900261,
-0.039511147886514664,
-0.016792481765151024,
-0.026050372049212456,
0.05515061691403389,
-0.026796113699674606,
0.0469350703060627,
0.010044436901807785,
0.08985920995473862,
-0.022402767091989517,
0.009400657378137112,
0.0371963307261467,
0.06908557564020157,
-0.0349... |
5def7663-5f9c-5bf5-88d5-3ac7b5815617 | /documentation/manage-data/collections/ | Collections | Create a Collection > Create Collection from Another Collection | documentation | true | Collections > Create a Collection > Create Collection from Another Collection
To create a collection from another collection, use the [Migration Tool](https://github.com/qdrant/migration/). You can use it to either copy a collection within the same Qdrant instance or to copy a collection to another instance.
For exam... | [
0.07179494947195053,
-0.059777550399303436,
-0.029236629605293274,
-0.06436414271593094,
0.05514729768037796,
-0.020469851791858673,
0.03528450056910515,
0.05700471252202988,
0.07558275759220123,
-0.04186802729964256,
0.022912684828042984,
0.0379939042031765,
0.03239980712532997,
-0.020759... |
5defe05a-d00c-5369-93f4-3d187ce721c2 | /course/multi-vector-search/module-2/colpali-family/ | ColPali Family Overview | Future of Multi-Vector Representations | course | false | ColPali Family Overview > Future of Multi-Vector Representations
The multi-vector approach extends naturally to new modalities beyond images. Models like **[TomoroAI/tomoro-colqwen3-embed-8b](https://huggingface.co/TomoroAI/tomoro-colqwen3-embed-8b)**, based on ColQwen3, demonstrate this extensibility by adding suppor... | [
0.03251116722822189,
-0.0412837378680706,
-0.02091536857187748,
-0.053184330463409424,
0.09104406833648682,
-0.020180528983473778,
0.06690413504838943,
0.02947281301021576,
0.029625430703163147,
-0.023360971361398697,
0.027263358235359192,
0.012539995834231377,
0.06762583553791046,
-0.0308... |
5dff5053-a78c-5775-90fd-9d17f8bb265c | /documentation/cloud-user-profile/ | User Profile | Invitations | documentation | false | User Profile > Invitations
When someone invites you to their account, the invitation appears on your **Invitations** page as a pending invitation, where you can accept it. Once accepted, the account becomes available in your [account switcher](/documentation/cloud-account-setup/#switching-between-accounts).
 is an open-source platform for authoring, scheduling and monitoring data and computing workflows. Airflow uses Python to create workflows that can be easily scheduled and monitored.
Qdrant is available as a [provider](https://airflow.apache.org/docs/apache-... | [
-0.0034714327193796635,
-0.026476481929421425,
-0.030783651396632195,
-0.06343217939138412,
0.08472687005996704,
-0.014209810644388199,
0.04563729837536812,
0.04707679525017738,
0.03858872875571251,
-0.007619538810104132,
0.028758574277162552,
-0.011503512039780617,
0.032376136630773544,
-... | |
5edf1c4d-39d5-5812-9a83-a7aa4b314d2b | /course/essentials/day-2/collection-tuning-demo/ | Demo: HNSW Performance Tuning | Step 8: Baseline Performance Testing | course | true | Demo: HNSW Performance Tuning > Step 8: Baseline Performance Testing
Let's measure search performance on the HNSW‑enabled collection.
```python
print("Running baseline performance test...")
# Warm up the RAM index/vectors cache with a test query
client.query_points(collection_name=collection_name, query=query_embedd... | [
0.010879794135689735,
-0.03503425419330597,
-0.0643189400434494,
-0.019622301682829857,
0.030539344996213913,
-0.009670726023614407,
0.04203767329454422,
0.0006667286506853998,
0.0865357518196106,
-0.00312711950391531,
0.03775876760482788,
0.04945537820458412,
0.02814112789928913,
-0.02926... |
5f067c59-35d1-5618-9a9e-cfd643ea77dc | /articles/discovery-search/ | Discovery Search in Qdrant | What is discovery search? | article | false | Discovery Search in Qdrant > What is discovery search?
In version 1.7, Qdrant [released](https://github.com/qdrant/qdrant/releases/tag/v1.7.0) this novel API that lets you constrain the space in which a search is performed, relying only on pure vectors. This is a powerful tool that lets you explore the vector space in... | [
0.05164455622434616,
-0.05309492349624634,
-0.047557711601257324,
0.0047394162975251675,
0.07088291645050049,
-0.005060246214270592,
0.05563117936253548,
0.03560255095362663,
0.05798868089914322,
-0.046084824949502945,
0.02633574604988098,
0.04382216930389404,
0.07204517722129822,
-0.01377... |
5f1e46fc-3bbb-51c5-9482-60deea530b0a | /documentation/tutorials-operations/incremental-embedding-updates/ | Incremental Embedding Updates | Run and Verify the Sync | documentation | true | Incremental Embedding Updates > Run and Verify the Sync
The five cases, assembled from the functions defined above:
```python
def sync(latest_chunks):
check_gate() # refuse to mix embedding models or pipeline versions
chunks = prepare_chunks_for_sync(latest_chunks)
incoming_ids, unchanged, content_chang... | [
0.06988085061311722,
-0.03776060417294502,
-0.03806162625551224,
-0.054311931133270264,
0.05672086402773857,
-0.02208658680319786,
0.009322872385382652,
0.011314923875033855,
0.06913814693689346,
-0.004734345246106386,
0.003247179090976715,
0.03143339604139328,
0.048198506236076355,
-0.038... |
5f201e16-3d2a-57c5-b843-00c271d14aff | /documentation/search/low-latency-search/ | Low-Latency Search | Create Payload Indexes | documentation | false | Low-Latency Search > Create Payload Indexes
If your search queries include filters, create [payload indexes](/documentation/manage-data/indexing/#payload-index) for the fields you filter on. Payload indexes are the primary way to improve filtered search performance in Qdrant. For best results, create payload indexes *... | [
0.0336994007229805,
-0.023723995313048363,
-0.05569702014327049,
-0.039024461060762405,
0.058319829404354095,
-0.04336013272404671,
0.04234476760029793,
0.015489996410906315,
0.07875434309244156,
-0.041897308081388474,
0.04054177179932594,
0.008731522597372532,
0.045165468007326126,
-0.011... |
5f224440-5880-59fb-8dac-ff61203a9419 | /course/essentials/day-1/chunking-strategies/ | Text Chunking Strategies | Chunking Strategies: The Shape Matters > 6. Semantic-Aware Chunking | course | true | Text Chunking Strategies > Chunking Strategies: The Shape Matters > 6. Semantic-Aware Chunking
**The Approach:** Use embeddings to detect meaning shifts and break at topic boundaries.
Everything up to this point has been about structure. But structure isn't the same as meaning. Semantic chunking uses embeddings to fi... | [
0.031449586153030396,
-0.02562631480395794,
-0.06806358695030212,
-0.05289017781615257,
0.043828047811985016,
-0.0327976755797863,
0.0672326385974884,
0.02026827074587345,
0.047810148447752,
-0.002991053508594632,
0.03348597139120102,
-0.005018246825784445,
0.05779437720775604,
-0.02095939... |
5f385d98-7a3c-56bb-943f-32125c6ace85 | /documentation/search/text-search/sparse-vector-search/ | Sparse Vector Search | BM25 > BM25 Text Processing > Tokenizer | documentation | true | Sparse Vector Search > BM25 > BM25 Text Processing > Tokenizer
The tokenizer breaks down text into individual tokens (words). By default, the sparse vector BM25 model uses the `word` tokenizer, which splits text based on word boundaries like whitespace and punctuation. This method is effective for Latin-based language... | [
0.03803715482354164,
-0.04552643373608589,
-0.004160525742918253,
-0.07028091698884964,
0.07349129766225815,
-0.039140623062849045,
0.055519454181194305,
0.016170157119631767,
0.015644598752260208,
-0.00473977392539382,
0.03424898162484169,
0.029488658532500267,
0.040283381938934326,
-0.01... |
5f50f6e8-ccf9-5e74-b41a-14cdc9f4a5f7 | /documentation/tutorials-search-engineering/using-relevance-feedback/ | Relevance Feedback Retrieval in Qdrant | Customizing for Your Use Case > Setup > Retriever | documentation | true | Relevance Feedback Retrieval in Qdrant > Customizing for Your Use Case > Setup > Retriever
The **retriever** is an embedding model that converts your raw data into vectors for semantic similarity search. The Relevance Feedback Query will help you optimize your retriever's ability to find relevant results in your data.... | [
0.04884593188762665,
-0.03750507906079292,
-0.03036932274699211,
-0.05442233383655548,
0.048275455832481384,
-0.022769005969166756,
0.06989552825689316,
0.04406372085213661,
0.05816393345594406,
-0.00966150313615799,
0.018845034763216972,
0.024977151304483414,
0.06715366989374161,
-0.04279... |
5f544a91-5077-5eff-bd68-8aa2f91c3f1f | /articles/dataset-quality/ | Detecting Dataset Errors with Similarity Search | Furniture web-marketplace > Similarity search | article | false | Detecting Dataset Errors with Similarity Search > Furniture web-marketplace > Similarity search
**Keep in mind the limits of this approach:**
- Distance only ranks the items: the furthest ones are candidates for a review, not confirmed errors. There is no universal threshold that tells where the errors end.
- The anch... | [
0.06753092259168625,
-0.01983199082314968,
-0.058509472757577896,
-0.009330498054623604,
0.057557158172130585,
-0.01196408923715353,
0.07086347043514252,
0.010118038393557072,
0.059190601110458374,
0.008734009228646755,
0.032357532531023026,
0.04048313945531845,
0.07082153856754303,
0.0123... |
5f686f26-442a-5efe-b297-4371f0b948e9 | /documentation/production-operations/vector-search-in-production/ | Vector Search in Production: Scaling, HA & Tuning Guide | 3. What's the Best Way to Scale the Database and Optimize Resources? > Preparing for High Availability Scenarios in Production | documentation | true | Vector Search in Production: Scaling, HA & Tuning Guide > 3. What's the Best Way to Scale the Database and Optimize Resources? > Preparing for High Availability Scenarios in Production
**Use at least 3 nodes** to ensure failover and reduce downtime risk.
A three-node setup provides a baseline for fault tolerance: if ... | [
-0.0005216380232013762,
-0.03081025369465351,
-0.07883066684007645,
-0.019232267513871193,
0.05164697393774986,
-0.011314359493553638,
0.045451853424310684,
0.02041393518447876,
0.05896533280611038,
-0.010790342465043068,
0.025277473032474518,
-0.02039434015750885,
0.05520400032401085,
-0.... |
5f6cfa1f-b28c-5794-9d4c-19b67b96b6e1 | /documentation/tutorials-build-essentials/video-anomaly-edge-part-1/ | Video Anomaly Detection Part 1: Architecture, Twelve Labs, and NVIDIA VSS | Learning Objectives | documentation | false | Video Anomaly Detection Part 1: Architecture, Twelve Labs, and NVIDIA VSS > Learning Objectives
In this series you will:
- **Build a kNN anomaly detector** using Qdrant vector search that detects novel anomaly types without per-category training.
- **Integrate Twelve Labs Marengo** for video embeddings that serve du... | [
0.05799601972103119,
-0.035982538014650345,
-0.08464164286851883,
-0.020517969503998756,
0.084117591381073,
-0.017398450523614883,
0.04278340935707092,
0.014020067639648914,
0.08652330935001373,
-0.026397408917546272,
0.07407716661691666,
0.006109949667006731,
0.08427438884973526,
-0.03757... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.