How to switch between different AI providers (OpenAI, Claude, etc.) in OpenClaw?

Clawpedia · For Humans

Change your AI provider in OpenClaw with a simple configuration update, supporting OpenAI, Anthropic, and more.

How to Switch Between Different AI Providers in OpenClaw

One of OpenClaw's greatest strengths is model agnosticism — you can switch between OpenAI, Anthropic, Google, local models, and dozens of other providers with a simple configuration change. No code modifications, no data migration, no downtime.

Quick Switch via CLI


# Check current model
openclaw config get model
# provider: openai
# model: gpt-4o

# Switch to Anthropic Claude
openclaw config set model.provider anthropic
openclaw config set model.model claude-3.5-sonnet
openclaw config set model.api_key $ANTHROPIC_API_KEY

# Switch to local Ollama
openclaw config set model.provider ollama
openclaw config set model.model llama3.1:8b

# Test the new model
openclaw chat "Hello, which model am I talking to?"

Provider Configuration Reference


# OpenAI
model:
  provider: openai
  model: gpt-4o              # gpt-4o, gpt-4, gpt-3.5-turbo
  api_key: ${OPENAI_API_KEY}
  endpoint: https://api.openai.com/v1  # default

# Anthropic
model:
  provider: anthropic
  model: claude-3.5-sonnet    # claude-4, claude-3-haiku
  api_key: ${ANTHROPIC_API_KEY}

# Google Gemini
model:
  provider: google
  model: gemini-2.5-pro       # gemini-2.5-flash
  api_key: ${GOOGLE_API_KEY}

# Ollama (Local)
model:
  provider: ollama
  model: llama3.1:70b         # Any Ollama model
  endpoint: http://localhost:11434

# OpenRouter (100+ models)
model:
  provider: openrouter
  model: anthropic/claude-3.5-sonnet
  api_key: ${OPENROUTER_API_KEY}

# Groq (Fast inference)
model:
  provider: groq
  model: llama-3.1-70b-versatile
  api_key: ${GROQ_API_KEY}

# Together AI
model:
  provider: together
  model: meta-llama/Llama-3.1-70B-Instruct
  api_key: ${TOGETHER_API_KEY}

# LM Studio (Local)
model:
  provider: lmstudio
  endpoint: http://localhost:1234/v1

Provider Comparison

ProviderBest ModelsSpeedCostPrivacyOffline
OpenAIGPT-4oFast$$Cloud❌
AnthropicClaude 3.5 SonnetFast$$Cloud❌
GoogleGemini 2.5 ProFast$$Cloud❌
OllamaLlama 3.1, MistralVariesFree✅ Local✅
GroqLlama 3.1 (fast)Ultra-fastFree tierCloud❌
OpenRouterAny modelVariesVariesCloud❌

Setting Up API Keys


# Store keys securely
openclaw secret set OPENAI_API_KEY
openclaw secret set ANTHROPIC_API_KEY
openclaw secret set GOOGLE_API_KEY

# List stored keys
openclaw secret list
# OPENAI_API_KEY      ****...7kQ2
# ANTHROPIC_API_KEY   ****...xP9m
# GOOGLE_API_KEY      ****...nR4w

Model Routing: Use Multiple Providers

LM StudioAny GGUFVariesFree✅ Local✅

Configure OpenClaw to use different models for different tasks:


model:
  default:
    provider: ollama
    model: llama3.1:8b           # Free, fast, handles most queries
  
  routes:
    - name: complex_reasoning
      provider: openai
      model: gpt-4o
      trigger: "needs deep analysis or complex reasoning"
    
    - name: code_generation
      provider: anthropic
      model: claude-3.5-sonnet
      trigger: "code writing or review"
    
    - name: quick_response
      provider: groq
      model: llama-3.1-8b-instant
      trigger: "simple factual questions"

Migration Between Providers

Switching providers has no impact on:

The only thing that changes is which AI model generates responses.

Troubleshooting Provider Issues

ProblemSolution
"Invalid API key"Verify key: openclaw secret get PROVIDER_API_KEY
"Model not found"Check exact model name (case-sensitive)
Slow responsesTry a faster provider or smaller model
Rate limit errorsSwitch to a provider with higher limits
"Connection refused" (Ollama)Start Ollama: ollama serve
Response quality droppedSome models are weaker — try a larger variant

Tip: Keep API keys for at least two providers configured. If one goes down or hits rate limits, you can switch in seconds with openclaw config set model.provider <backup>.

Related Articles