LLM Providers - OpenAnalyst CLI Docs

Supported Providers

Recommended: OpenAnalyst account

Run openanalyst account login for OTP-based sign-in. Create a free account at 10x.in to get free credits plus access to the skills and plugins catalogs. For other providers, bring your own API key via env vars.

OpenAnalyst Default

OpenAnalyst's live routed catalog with OpenAnalyst, GPT, Claude, Gemini, DeepSeek, Qwen, and free models fetched live.

openanalyst account login or OPENANALYST_API_KEY

Free account + credits at 10x.in.

Anthropic / Claude

Claude Opus 4.6, Sonnet 4.6, Haiku with extended thinking support.

ANTHROPIC_API_KEY

OpenAI / Codex

GPT-5.4, GPT-4.1, GPT-4o, o3, o4 Mini, and Codex via direct OpenAI API.

OPENAI_API_KEY

Google Gemini

Gemini 3.1 Pro, Gemini 3.1 Flash with live model discovery.

GEMINI_API_KEY

xAI / Grok

Grok 4, Grok Mini with streaming tool support.

XAI_API_KEY

OpenRouter

350+ models from any provider through a single gateway.

OPENROUTER_API_KEY

Stability AI

Stable Diffusion and image generation models.

STABILITY_API_KEY

Amazon Bedrock

AWS-hosted models with live discovery from your Bedrock account.

BEDROCK_API_KEY

Mid-Conversation Model Switching

One of OpenAnalyst's defining features: switch models without losing your conversation context.

# Start with Claude
openanalyst --model opus
> explain this codebase

# Switch to GPT-4o mid-conversation
/model gpt-4o
> now refactor the auth module

# Switch to Gemini for review
/model gemini-3.1-pro
> review the changes for security issues

How It Works

The session state (messages, tool results, system prompt) is provider-agnostic. When you switch models, the runtime translates the conversation into the new provider's format while preserving full context.

Model Aliases

Use short aliases instead of full model IDs:

Alias	Resolves To	Provider
`opus`	claude-opus-4-6	Anthropic
`sonnet`	claude-sonnet-4-6	Anthropic
`haiku`	claude-haiku-4-5	Anthropic
`gpt-5`	gpt-5.4	OpenAI
`gpt-4o`	gpt-4o	OpenAI
`o3`	o3	OpenAI
`codex-mini`	codex-mini-latest	OpenAI
`gemini-pro`	gemini-3.1-pro	Google
`gemini-flash`	gemini-3.1-flash	Google
`grok`	grok-4	xAI
`grok-mini`	grok-4-mini	xAI
`deepseek`	deepseek-v4	OpenRouter
`llama`	llama-4-maverick	OpenRouter

Live Model Discovery

All providers support live model discovery — available models are fetched from the provider's API or OpenAnalyst routing layer at runtime, so you always have access to the latest models without updating the CLI.

# List available models for current provider
/model

# List models from a specific provider
openanalyst models --provider openai

Provider Feature Matrix

Feature	OpenAnalyst	Anthropic	OpenAI	Gemini	xAI	Stability	OpenRouter	Bedrock
Streaming	✓	✓	✓	✓	✓	—	✓	✓
Tool calling	✓	✓	✓	✓	✓	—	✓	✓
Live discovery	✓	✓	✓	✓	✓	✓	✓	✓
Extended thinking	✓	✓	—	—	—	—	Varies	—
Image generation	—	—	✓	✓	—	✓	Varies	—
Interactive login	✓	✓	✓	✓	✓	✓	✓	✓