File size: 5,243 Bytes
aadaf89 f29b29e aadaf89 f29b29e be81b00 f29b29e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 | ---
title: README
emoji: ⚡
colorFrom: yellow
colorTo: gray
sdk: static
pinned: false
---
<p align="center">
<a href="https://dakera.ai"><img src="https://dakera.ai/assets/logo-128.png" width="104" alt="Dakera logo"></a>
</p>
<h1 align="center">Dakera</h1>
<p align="center"><b>Self-hosted memory engine for AI agents</b><br>Long-term memory that survives sessions, in one Rust binary on your own servers.</p>
<p align="center">
<a href="https://dakera.ai"><img alt="Website" src="https://img.shields.io/badge/Website-dakera.ai-c9a227?style=flat-square"></a>
<a href="https://dakera.ai/docs"><img alt="Docs" src="https://img.shields.io/badge/Docs-dakera.ai%2Fdocs-0b0b0b?style=flat-square"></a>
<a href="https://github.com/Dakera-AI"><img alt="GitHub" src="https://img.shields.io/badge/GitHub-Dakera--AI-181717?style=flat-square&logo=github&logoColor=white"></a>
<a href="https://www.linkedin.com/company/dakera-ai"><img alt="LinkedIn" src="https://img.shields.io/badge/LinkedIn-Dakera%20AI-0a66c2?style=flat-square"></a>
</p>
**Dakera gives AI agents a memory.** Store what your agents learn, recall it by meaning and by keyword, and let it fade or strengthen with use: hybrid vector and full-text retrieval, cross-encoder reranking, entity extraction, sessions and importance decay, behind a REST, gRPC and MCP API. Everything runs on your own infrastructure, CPU-only if you like, with no external AI service.
```bash
docker run -p 3000:3000 -e DAKERA_ROOT_API_KEY=dk-mykey ghcr.io/dakera-ai/dakera:latest
```
## Models
Dakera runs its models locally with ONNX Runtime. This organisation holds verified copies of every model Dakera downloads: byte-for-byte identical to the originals, credited to their authors under their own licences, and checked against the same SHA-256 whichever source a server downloads from.
| Role in Dakera | Models |
|---|---|
| 🧠 Embedding | [bge-large-en-v1.5](https://huggingface.co/DakeraAI/bge-large-en-v1.5) (default) · [bge-small-en-v1.5](https://huggingface.co/DakeraAI/bge-small-en-v1.5) · [e5-small-v2](https://huggingface.co/DakeraAI/e5-small-v2) · [all-MiniLM-L6-v2](https://huggingface.co/DakeraAI/all-MiniLM-L6-v2) · [bge-m3](https://huggingface.co/DakeraAI/bge-m3) · [modernbert-embed-base](https://huggingface.co/DakeraAI/modernbert-embed-base) · [gte-modernbert-base](https://huggingface.co/DakeraAI/gte-modernbert-base) |
| 🎯 Late interaction | [answerai-colbert-small-v1](https://huggingface.co/DakeraAI/answerai-colbert-small-v1) |
| 🏆 Reranking | [bge-reranker-v2-m3-ONNX](https://huggingface.co/DakeraAI/bge-reranker-v2-m3-ONNX) (default) · [bge-reranker-base](https://huggingface.co/DakeraAI/bge-reranker-base) |
| 🏷️ Entity extraction (int8, built by Dakera) | [gliner-bi-small](https://huggingface.co/DakeraAI/gliner-bi-small-v2.0-int8) (default, English) · [gliner-x-small](https://huggingface.co/DakeraAI/gliner-x-small-int8) (default, multilingual) · [gliner-bi-edge](https://huggingface.co/DakeraAI/gliner-bi-edge-v2.0-int8) · [gliner-medium](https://huggingface.co/DakeraAI/gliner_medium-v2.1-int8) · [gliner-multi](https://huggingface.co/DakeraAI/gliner_multi-v2.1-int8) |
| 🗂️ Intent detection | [paraphrase-multilingual-MiniLM-L12-v2](https://huggingface.co/DakeraAI/paraphrase-multilingual-MiniLM-L12-v2) (forget requests, standing instructions) |
| 🎙️ Speech to text | [whisper-tiny.en](https://huggingface.co/DakeraAI/whisper-tiny.en) · [whisper-base.en](https://huggingface.co/DakeraAI/whisper-base.en) · [whisper-tiny](https://huggingface.co/DakeraAI/whisper-tiny) · [whisper-base](https://huggingface.co/DakeraAI/whisper-base) (default) · [whisper-small](https://huggingface.co/DakeraAI/whisper-small) |
| 🖼️ Image and page retrieval | [colmodernvbert](https://huggingface.co/DakeraAI/colmodernvbert) |
## Build with Dakera
| | |
|---|---|
| 🚀 Self-hosting: Docker, Kubernetes, Helm | [dakera-deploy](https://github.com/Dakera-AI/dakera-deploy) · [dakera-helm](https://github.com/Dakera-AI/dakera-helm) |
| 🔌 MCP server for Claude, Cursor and other agents | [dakera-mcp](https://github.com/Dakera-AI/dakera-mcp) |
| 🧰 SDKs | [Python](https://github.com/Dakera-AI/dakera-py) · [TypeScript](https://github.com/Dakera-AI/dakera-js) · [Go](https://github.com/Dakera-AI/dakera-go) · [Rust](https://github.com/Dakera-AI/dakera-rs) |
| 🦜 Frameworks | [LangChain](https://github.com/Dakera-AI/dakera-langchain) · [LangChain.js](https://github.com/Dakera-AI/dakera-langchain-js) · [LlamaIndex](https://github.com/Dakera-AI/dakera-llamaindex) · [CrewAI](https://github.com/Dakera-AI/dakera-crewai) · [AutoGen](https://github.com/Dakera-AI/dakera-autogen) · [Vercel AI SDK](https://github.com/Dakera-AI/dakera-ai-sdk) · [Strands](https://github.com/Dakera-AI/strands-dakera) |
| ⌨️ Command line | [dakera-cli](https://github.com/Dakera-AI/dakera-cli) |
## Get in touch
Questions, feedback or a feature you need: **[amine@dakera.ai](mailto:amine@dakera.ai)**.
<p align="center"><a href="https://dakera.ai">dakera.ai</a> · <a href="https://dakera.ai/docs">Docs</a> · <a href="https://github.com/Dakera-AI">GitHub</a> · <a href="https://www.linkedin.com/company/dakera-ai">LinkedIn</a></p>
|