docs: add OpenClaw integration tutorial

This commit is contained in:
shin-bot-litellm 2026-01-31 05:09:51 +00:00
parent 3f1bda57e2
commit 1895bded68

View file

@ -0,0 +1,162 @@
# Using LiteLLM with OpenClaw
[OpenClaw](https://openclaw.ai) is a personal AI assistant that runs locally and connects to your messaging platforms (Slack, Discord, Telegram, etc.). LiteLLM can be used as the model provider backend, giving you access to 100+ LLMs through a unified API.
## Why use LiteLLM with OpenClaw?
- **Model flexibility**: Switch between OpenAI, Anthropic, Azure, Bedrock, and more without changing your OpenClaw config
- **Cost tracking**: Monitor spend across all your AI interactions
- **Enterprise support**: Use Azure OpenAI, AWS Bedrock, or Vertex AI through LiteLLM proxy
- **Load balancing**: Distribute requests across multiple API keys or providers
## Quick Start
### 1. Start LiteLLM Proxy
```bash
litellm --model gpt-4o
```
Or with a config file for multiple models:
```yaml
# litellm_config.yaml
model_list:
- model_name: gpt-4o
litellm_params:
model: gpt-4o
api_key: os.environ/OPENAI_API_KEY
- model_name: claude-sonnet
litellm_params:
model: claude-sonnet-4-20250514
api_key: os.environ/ANTHROPIC_API_KEY
```
```bash
litellm --config litellm_config.yaml
```
### 2. Configure OpenClaw
In your OpenClaw config (`~/.openclaw/config.yaml`):
```yaml
providers:
litellm:
baseUrl: http://localhost:4000/v1
apiKey: sk-your-litellm-key # Optional if no auth configured
models:
default: litellm/gpt-4o
# Or use Claude through LiteLLM:
# default: litellm/claude-sonnet
```
### 3. Start OpenClaw
```bash
openclaw gateway start
```
OpenClaw will now route all model requests through your LiteLLM proxy.
## Using LiteLLM Cloud (Hosted)
If you're using [LiteLLM Cloud](https://litellm.ai), configure OpenClaw with your cloud endpoint:
```yaml
providers:
litellm:
baseUrl: https://api.litellm.ai/v1
apiKey: sk-your-litellm-cloud-key
models:
default: litellm/gpt-4o
```
## Advanced Configuration
### Multiple Model Routing
Configure different models for different tasks:
```yaml
providers:
litellm:
baseUrl: http://localhost:4000/v1
models:
default: litellm/gpt-4o
reasoning: litellm/o1-preview
fast: litellm/gpt-4o-mini
vision: litellm/gpt-4o
```
### Using Azure OpenAI through LiteLLM
LiteLLM config:
```yaml
model_list:
- model_name: gpt-4o
litellm_params:
model: azure/gpt-4o-deployment
api_base: https://your-resource.openai.azure.com
api_key: os.environ/AZURE_API_KEY
api_version: "2024-02-15-preview"
```
OpenClaw sees this as `litellm/gpt-4o` - no Azure-specific config needed.
### Using AWS Bedrock through LiteLLM
LiteLLM config:
```yaml
model_list:
- model_name: claude-sonnet
litellm_params:
model: bedrock/anthropic.claude-sonnet-4-20250514-v1:0
aws_access_key_id: os.environ/AWS_ACCESS_KEY_ID
aws_secret_access_key: os.environ/AWS_SECRET_ACCESS_KEY
aws_region_name: us-east-1
```
OpenClaw config:
```yaml
models:
default: litellm/claude-sonnet
```
## Tool Calling & Function Support
LiteLLM fully supports tool/function calling, which OpenClaw uses extensively. All tools defined in your OpenClaw skills will work through LiteLLM.
Monitor tool calls in the LiteLLM dashboard:
- See which tools were called per request
- Track success/failure rates
- Debug tool execution issues
## Cost Tracking
LiteLLM tracks costs for all requests. View spend in the LiteLLM UI or via API:
```bash
curl http://localhost:4000/spend/logs
```
## Troubleshooting
### Connection refused
Ensure LiteLLM proxy is running and the `baseUrl` is correct.
### Model not found
Check that the model name in OpenClaw matches a `model_name` in your LiteLLM config.
### Authentication errors
Verify your API keys are set correctly in the LiteLLM config.
## Resources
- [LiteLLM Documentation](https://docs.litellm.ai)
- [OpenClaw Documentation](https://docs.openclaw.ai)
- [LiteLLM GitHub](https://github.com/BerriAI/litellm)
- [OpenClaw GitHub](https://github.com/openclaw/openclaw)