Set up cognee integrations
All integration config is environment variables (.env). The authoritative,
always-current list with commented examples is .env.template at the repo
root — check it before inventing variable names. Install the matching extra
before switching a backend (e.g. pip install cognee[postgres]).
LLM providers
Default is OpenAI (LLM_API_KEY is all you need). To switch, set
LLM_PROVIDER, LLM_MODEL, LLM_API_KEY, and (where relevant)
LLM_ENDPOINT / LLM_API_VERSION:
- Azure OpenAI:
LLM_PROVIDER=azure, LLM_MODEL=azure/gpt-4o-mini, endpoint + api version required.
- Gemini (no extra needed):
LLM_PROVIDER=gemini, LLM_MODEL=gemini/gemini-2.0-flash-exp.
- Anthropic (
cognee[anthropic]): LLM_PROVIDER=anthropic, model e.g. claude-3-5-sonnet-20241022.
- Ollama, local (
cognee[ollama]): LLM_PROVIDER=ollama, LLM_ENDPOINT=http://localhost:11434/v1, and set the embedding block + HUGGINGFACE_TOKENIZER too.
- Custom / OpenRouter / vLLM:
LLM_PROVIDER=custom with the provider's OpenAI-compatible endpoint.
- AWS Bedrock (
cognee[aws]): LLM_PROVIDER=bedrock + AWS credentials/region.
The classic trap: LLM and embeddings are configured independently
(EMBEDDING_PROVIDER, EMBEDDING_MODEL, EMBEDDING_ENDPOINT,
EMBEDDING_API_KEY). Configuring only one leaves the other on OpenAI —
either keep a valid OpenAI key or configure both.
Databases
- Relational (
DB_PROVIDER): sqlite (default) or postgres
(cognee[postgres]; host/port/user/password/name via DB_* vars).
- Vector (
VECTOR_DB_PROVIDER): lancedb (default), pgvector
(cognee[postgres], needs VECTOR_DB_URL), neptune_analytics
(cognee[neptune]), turso (cognee[turso]). Anything else (ChromaDB,
Qdrant, Weaviate, Milvus, …) lives in community adapters — install from
https://github.com/topoteretes/cognee-community and register with
use_vector_adapter before use; setting VECTOR_DB_PROVIDER alone raises
"Unsupported vector database provider".
- Graph (
GRAPH_DATABASE_PROVIDER): ladybug (default), neo4j
(cognee[neo4j], bolt URL + credentials), neptune (cognee[neptune]),
ladybug-remote, postgres (no raw Cypher / natural-language search).
The repo docker-compose.yml ships ready-to-use postgres (pgvector) and
neo4j profiles with matching default credentials. From a container, reach
host services with DB_HOST=host.docker.internal.
Storage, cache, and the rest
- S3 storage (
cognee[aws]): STORAGE_BACKEND=s3 + bucket/credentials,
and point DATA_ROOT_DIRECTORY/SYSTEM_ROOT_DIRECTORY at s3:// paths.
- Session cache:
CACHE_BACKEND = sqlite (default) | postgres | redis | fs | tapes.
- Ontologies:
ONTOLOGY_FILE_PATH to an OWL file, resolver/matching via
ONTOLOGY_RESOLVER / MATCHING_STRATEGY.
MCP server (IDE integration)
docker compose --profile mcp up starts the MCP server on port 8001
(SSE transport), built from cognee-mcp/. Point Cursor / Claude Desktop /
Claude Code at it to use cognee memory from the IDE. Configure its DB_* env
to match the main service so both see the same data.
After changing providers mid-project
Embeddings from different models are not comparable — after switching the
embedding provider or model, reset local state (cognee-cli forget --all or
await cognee.forget(everything=True)) and re-ingest with remember().
To drop just the graph and vectors while keeping the ingested files, use
await cognee.forget(dataset="my_project", memory_only=True) — the dataset can
then be rebuilt under the new embedding model without re-uploading anything.
1---2name: cognee-integrations3description: Use when the user wants to connect cognee to external services — switching LLM or embedding providers (OpenAI, Azure, Gemini, Anthropic, Ollama, OpenRouter), changing databases (Postgres, PGVector, Neo4j, Neptune, Turso), S3 storage, or the MCP server for IDE integration.4---5
6# Set up cognee integrations
7
8All integration config is environment variables (`.env`). The authoritative,
9always-current list with commented examples is `.env.template` at the repo
10root — check it before inventing variable names. Install the matching extra
11before switching a backend (e.g. `pip install cognee[postgres]`).
12
13## LLM providers
14
15Default is OpenAI (`LLM_API_KEY` is all you need). To switch, set
16`LLM_PROVIDER`, `LLM_MODEL`, `LLM_API_KEY`, and (where relevant)
17`LLM_ENDPOINT` / `LLM_API_VERSION`:
18
19- **Azure OpenAI**: `LLM_PROVIDER=azure`, `LLM_MODEL=azure/gpt-4o-mini`, endpoint + api version required.
20- **Gemini** (no extra needed): `LLM_PROVIDER=gemini`, `LLM_MODEL=gemini/gemini-2.0-flash-exp`.
21- **Anthropic** (`cognee[anthropic]`): `LLM_PROVIDER=anthropic`, model e.g. `claude-3-5-sonnet-20241022`.
22- **Ollama, local** (`cognee[ollama]`): `LLM_PROVIDER=ollama`, `LLM_ENDPOINT=http://localhost:11434/v1`, and set the embedding block + `HUGGINGFACE_TOKENIZER` too.
23- **Custom / OpenRouter / vLLM**: `LLM_PROVIDER=custom` with the provider's OpenAI-compatible endpoint.
24- **AWS Bedrock** (`cognee[aws]`): `LLM_PROVIDER=bedrock` + AWS credentials/region.
25
26**The classic trap**: LLM and embeddings are configured independently
27(`EMBEDDING_PROVIDER`, `EMBEDDING_MODEL`, `EMBEDDING_ENDPOINT`,
28`EMBEDDING_API_KEY`). Configuring only one leaves the other on OpenAI —
29either keep a valid OpenAI key or configure both.
30
31## Databases
32
33- **Relational** (`DB_PROVIDER`): sqlite (default) or postgres
34 (`cognee[postgres]`; host/port/user/password/name via `DB_*` vars).
35- **Vector** (`VECTOR_DB_PROVIDER`): lancedb (default), pgvector
36 (`cognee[postgres]`, needs `VECTOR_DB_URL`), neptune_analytics
37 (`cognee[neptune]`), turso (`cognee[turso]`). Anything else (ChromaDB,
38 Qdrant, Weaviate, Milvus, …) lives in community adapters — install from
39 https://github.com/topoteretes/cognee-community and register with
40 `use_vector_adapter` before use; setting `VECTOR_DB_PROVIDER` alone raises
41 "Unsupported vector database provider".
42- **Graph** (`GRAPH_DATABASE_PROVIDER`): ladybug (default), neo4j
43 (`cognee[neo4j]`, bolt URL + credentials), neptune (`cognee[neptune]`),
44 ladybug-remote, postgres (no raw Cypher / natural-language search).
45
46The repo `docker-compose.yml` ships ready-to-use `postgres` (pgvector) and
47`neo4j` profiles with matching default credentials. From a container, reach
48host services with `DB_HOST=host.docker.internal`.
49
50## Storage, cache, and the rest
51
52- **S3 storage** (`cognee[aws]`): `STORAGE_BACKEND=s3` + bucket/credentials,
53 and point `DATA_ROOT_DIRECTORY`/`SYSTEM_ROOT_DIRECTORY` at `s3://` paths.
54- **Session cache**: `CACHE_BACKEND` = sqlite (default) | postgres | redis | fs | tapes.
55- **Ontologies**: `ONTOLOGY_FILE_PATH` to an OWL file, resolver/matching via
56 `ONTOLOGY_RESOLVER` / `MATCHING_STRATEGY`.
57
58## MCP server (IDE integration)
59
60`docker compose --profile mcp up` starts the MCP server on port 8001
61(SSE transport), built from `cognee-mcp/`. Point Cursor / Claude Desktop /
62Claude Code at it to use cognee memory from the IDE. Configure its `DB_*` env
63to match the main service so both see the same data.
64
65## After changing providers mid-project
66
67Embeddings from different models are not comparable — after switching the
68embedding provider or model, reset local state (`cognee-cli forget --all` or
69`await cognee.forget(everything=True)`) and re-ingest with `remember()`.
70
71To drop just the graph and vectors while keeping the ingested files, use
72`await cognee.forget(dataset="my_project", memory_only=True)` — the dataset can
73then be rebuilt under the new embedding model without re-uploading anything.
topoteretes