Models

Supported LLM Providers

Six first-party LLM providers (OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Qwen, Kimi) plus Azure OpenAI, Codex OAuth and OpenRouter — and any OpenAI-compatible endpoint you self-host. Mix providers per node to tune speed, cost and quality.

OpenAI

Typeopenai
NotesGPT-4, GPT-5 series
AuthAPI Key

Anthropic Claude

Typeclaude
NotesClaude 4 Sonnet / Opus
AuthAPI Key · Subscription Token

Google Gemini

Typegemini
NotesGemini 2.5+ Pro / Flash
AuthAPI Key

DeepSeek

Typedeepseek
NotesChat & Reasoning
AuthAPI Key

Alibaba Qwen

Typeopenai
NotesOpenAI-compatible
AuthAPI Key (OpenAI-compatible)

Moonshot Kimi

Typeopenai
Noteskimi-k2-turbo-preview
AuthAPI Key (OpenAI-compatible)

MiniMax

Typeminimax
NotesMiniMax models
AuthAPI Key

GLM (Zhipu)

Typeglm
NotesGLM-4 series
AuthAPI Key

Azure OpenAI

Typeopenai
NotesEnterprise deployments
AuthAzure Key + endpoint + deployment

OpenAI Codex

Typecodex
NotesOAuth, no API key
AuthOAuth (local Codex credential)

OpenRouter

Typeopenrouter
Notes300+ models, single key
AuthAPI Key (single key, 300+ models)
agent.ymlyaml
ProviderTypeNotes
OpenAIopenaiGPT-4, GPT-5 series
Anthropic ClaudeclaudeClaude 4 Sonnet / Opus
Google GeminigeminiGemini 2.5+ Pro / Flash
DeepSeekdeepseekChat & Reasoning
Alibaba QwenopenaiOpenAI-compatible
Moonshot Kimiopenaikimi-k2-turbo-preview
MiniMaxminimaxMiniMax models
GLM (Zhipu)glmGLM-4 series
Azure OpenAIopenaiEnterprise deployments
OpenAI CodexcodexOAuth, no API key
OpenRouteropenrouter300+ models, single key
Authentication

Authentication matrix

How each provider authenticates, and whether it works through Azure or a self-hosted OpenAI-compatible endpoint.

auth-matrix.ymlyaml
ProviderAuthenticationAzureSelf-Hosted
OpenAIAPI Key✔ via Azure OpenAI✔ any OpenAI-compatible
Anthropic ClaudeAPI Key · Subscription Token
Google GeminiAPI Key
DeepSeekAPI Key
Alibaba QwenAPI Key (OpenAI-compatible)
Moonshot KimiAPI Key (OpenAI-compatible)
MiniMaxAPI Key
GLM (Zhipu)API Key
Azure OpenAIAzure Key + endpoint + deployment✔ native
OpenAI CodexOAuth (local Codex credential)
OpenRouterAPI Key (single key, 300+ models)
Strategy

Per-Node strategy guide

Recommended provider tier for each workflow stage. Combine with the pipeline YAML below to wire the routing.

Schema Linking

Fast / cheap: DeepSeek Chat, gpt-4o-mini. Runs on every turn to pick relevant tables.

SQL Generation

Strong: Claude Sonnet, gpt-4-turbo. Writes the SQL that downstream steps validate.

Reasoning / Output

Strongest: Claude Opus, gpt-4-turbo. Explains results and drives multi-step reasoning.
Embeddings

Embedding Models for Context Recall

Vectorize text and queries for semantic search, retrieval, and context matching.

OpenAI Embeddings

text-embedding-3-*1536 / 3072 dim, hosted

sentence-transformers

all-MiniLM-L6-v2384 dim · ~100MB · default

Multilingual E5

e5-large-instruct1024 dim · ~1.2GB · multilingual

BGE (zh)

bge-large-zh-v1.51024 dim · Chinese optimized

BGE (en)

bge-large-en-v1.51024 dim · English optimized
Observability

LLM Observability

Trace every prompt in production or debug locally as YAML — pick the surface that fits your workflow.

LangSmith

PurposeLLM call tracing & debugging
EnvLANGSMITH_TRACING=true · LANGSMITH_API_KEY

Langfuse

PurposeAgent + tool full-chain traces (OTel / OpenInference)
EnvLANGFUSE_PUBLIC_KEY · LANGFUSE_SECRET_KEY · LANGFUSE_HOST

LLM Trace

PurposeLocal YAML dump of prompts & completions
Env--save_llm_trace
Per-node routing

Per-Node Model Assignment

Use different providers for different workflow steps. Route schema linking to a cheaper model, SQL generation to a stronger one — all within the same pipeline.

Schema linking → fast/cheapSQL generation → strongValidation → balanced
pipeline.ymlyaml
pipeline:
  schema_linking:
    provider: openai
    model: gpt-4o-mini
  sql_generation:
    provider: claude
    model: claude-sonnet-4
  validation:
    provider: gemini
    model: gemini-2.5-flash
FAQ

Frequently asked questions

Supported providers, per-node routing, switching models, embeddings, and tracing.

Which LLM providers does Datus support?

Datus supports 11+ providers out of the box: OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Alibaba Qwen, Moonshot Kimi, MiniMax, GLM (Zhipu), Azure OpenAI, OpenAI Codex via OAuth, and OpenRouter for 300+ models behind a single key.

Can I use a different model for each step of a workflow?

Yes. Per-node model assignment lets you pick a different provider and model for each step — for example, a fast model for planning, a strong reasoning model for SQL generation, and a cheap model for validation.

How do I switch the active model from the CLI?

Run /model inside the Datus CLI to list configured providers and pick a new active one. The change is written back to ~/.datus/agent.yml, so the next session keeps the same default.

What about embeddings?

Datus supports OpenAI embeddings, Sentence-Transformers, and Hugging Face models (E5, BGE), so you can keep retrieval entirely local or use a hosted provider as you prefer.

Do I need keys from every provider?

No — you only configure the providers you actually use. Many teams start with a single key (OpenAI or OpenRouter) and add specialized models later for cost or latency reasons.

How do I trace LLM calls?

Datus supports LangSmith, Langfuse (with OpenTelemetry / OpenInference) and a local --save_llm_trace flag that dumps every prompt and completion to YAML for offline inspection.

Route the Right Model at Every Step

Configure planners, coders and rerankers per provider — Anthropic, OpenAI, Gemini, or your own self-hosted endpoint, all from one YAML.