.env.example

# =============================================================================
# all-agentic-architectures — environment variables
# Copy this file to `.env` and fill in your own values.  .env is gitignored.
# =============================================================================

# ---------------------------------------------------------------------------
# 1.  LLM PROVIDER SELECTION
# ---------------------------------------------------------------------------
# Pick ONE provider — every notebook + every architecture in the library will
# use whichever provider you set here, with no code changes.
#
# Supported values:
#   nebius | openai | anthropic | groq | ollama | together | fireworks |
#   mistralai | google
#
LLM_PROVIDER=nebius

# Model name for the chosen provider. Sensible defaults per provider:
#   nebius     → meta-llama/Meta-Llama-3.1-70B-Instruct
#   openai     → gpt-4o-mini
#   anthropic  → claude-sonnet-4-5
#   groq       → llama-3.3-70b-versatile
#   ollama     → llama3.1:8b
#   together   → meta-llama/Llama-3.3-70B-Instruct-Turbo
#   google     → gemini-2.0-flash
LLM_MODEL=meta-llama/Llama-3.3-70B-Instruct

# Default sampling temperature (architectures override per-node when needed).
LLM_TEMPERATURE=0.0

# ---------------------------------------------------------------------------
# 2.  PROVIDER API KEYS — only fill in the one(s) you use
# ---------------------------------------------------------------------------
NEBIUS_API_KEY=
OPENAI_API_KEY=
ANTHROPIC_API_KEY=
GROQ_API_KEY=
TOGETHER_API_KEY=
FIREWORKS_API_KEY=
MISTRAL_API_KEY=
GOOGLE_API_KEY=

# Ollama runs locally; no key needed. Override the host if not on default port.
OLLAMA_BASE_URL=http://localhost:11434

# ---------------------------------------------------------------------------
# 3.  TOOL APIs
# ---------------------------------------------------------------------------
# Tavily web search — used by tool-using / ReAct / RAG architectures.
# Free tier: https://app.tavily.com  (1000 calls / month free)
TAVILY_API_KEY=

# ---------------------------------------------------------------------------
# 4.  LANGSMITH TRACING (highly recommended — free tier is generous)
# ---------------------------------------------------------------------------
# When set, every agent run is traced at https://smith.langchain.com
LANGSMITH_API_KEY=
LANGSMITH_PROJECT=all-agentic-architectures
LANGCHAIN_TRACING_V2=true

# ---------------------------------------------------------------------------
# 5.  VECTOR STORE (used by RAG + episodic memory notebooks)
# ---------------------------------------------------------------------------
# Choose backend: faiss (local, default) | chroma (local) | qdrant (server)
VECTOR_BACKEND=faiss

# Only needed if VECTOR_BACKEND=qdrant
QDRANT_URL=http://localhost:6333
QDRANT_API_KEY=

# ---------------------------------------------------------------------------
# 6.  GRAPH DATABASE (used by graph-memory + GraphRAG notebooks)
# ---------------------------------------------------------------------------
# `docker compose up neo4j` brings up a local instance matching these defaults.
NEO4J_URI=bolt://localhost:7687
NEO4J_USERNAME=neo4j
NEO4J_PASSWORD=changeme_local_only

# ---------------------------------------------------------------------------
# 7.  EMBEDDINGS PROVIDER (defaults to LLM_PROVIDER unless overridden)
# ---------------------------------------------------------------------------
# Leave blank to inherit from LLM_PROVIDER above, or set explicitly:
#   nebius (BAAI/bge-en-icl) | openai (text-embedding-3-small) |
#   huggingface (sentence-transformers/all-MiniLM-L6-v2) | ollama (nomic-embed-text)
EMBEDDINGS_PROVIDER=
EMBEDDINGS_MODEL=