Model Selection and Configuration¶
Elroy supports various AI models for chat and embedding functionality.
Automatic Model Selection¶
Elroy will automatically select appropriate models based on available API keys:
| API Key | Chat Model | Embedding Model |
|---|---|---|
ANTHROPIC_API_KEY |
Claude 4.5 Sonnet (claude-sonnet-4-5-20250929) |
text-embedding-3-small |
OPENAI_API_KEY |
GPT-5 (gpt-5) |
text-embedding-3-small |
GEMINI_API_KEY |
Gemini 2.0 Flash (gemini/gemini-2.0-flash) |
text-embedding-3-small |
Model Configuration Options¶
Set via environment variable or config file:
| Config Key | Env Variable | Description | Default |
|---|---|---|---|
chat_model |
ELROY_CHAT_MODEL |
The model to use for chat completions | Inferred from API keys |
chat_model_api_base |
ELROY_CHAT_MODEL_API_BASE |
Base URL for OpenAI compatible chat model API | - |
chat_model_api_key |
ELROY_CHAT_MODEL_API_KEY |
API key for OpenAI compatible chat model API | - |
embedding_model |
ELROY_EMBEDDING_MODEL |
The model to use for text embeddings | text-embedding-3-small |
embedding_model_size |
ELROY_EMBEDDING_MODEL_SIZE |
The size of the embedding model | 1536 |
embedding_model_api_base |
ELROY_EMBEDDING_MODEL_API_BASE |
Base URL for OpenAI compatible embedding model API | - |
embedding_model_api_key |
ELROY_EMBEDDING_MODEL_API_KEY |
API key for OpenAI compatible embedding model API | - |
enable_caching |
ELROY_ENABLE_CACHING |
Whether to enable caching for the LLM | true |
Model Aliases¶
The following aliases can be used as the value of chat_model (via ELROY_CHAT_MODEL or config file):
| Alias | Model |
|---|---|
sonnet |
Claude 4.5 Sonnet (claude-sonnet-4-5-20250929) |
opus |
Claude 4.5 Opus (claude-opus-4-5-20251101) |
haiku |
Claude 3.5 Haiku (claude-3-5-haiku-20241022) |
o1 |
OpenAI o1 (openai/o1) |
Example:
Or in config file: