URL Format
The proxy URL follows a simple, predictable pattern.
Structure
https://proxy.aispendops.com/v1/{provider}/{provider-api-path}
| Segment | Description |
|---|---|
proxy.aispendops.com | Proxy host |
/v1/ | API version prefix |
{provider} | Provider identifier (e.g. openai, anthropic) |
{provider-api-path} | Everything after the provider is forwarded as-is |
Examples
| Provider | Proxy URL | Upstream |
|---|---|---|
| OpenAI | https://proxy.aispendops.com/v1/openai/v1/chat/completions | https://api.openai.com/v1/chat/completions |
| Anthropic | https://proxy.aispendops.com/v1/anthropic/v1/messages | https://api.anthropic.com/v1/messages |
https://proxy.aispendops.com/v1/google-ai-studio/v1beta/openai/chat/completions | https://generativelanguage.googleapis.com/v1beta/openai/chat/completions | |
| Groq | https://proxy.aispendops.com/v1/groq/v1/chat/completions | https://api.groq.com/openai/v1/chat/completions |
Any endpoint works
The proxy routes based on the path and does not interpret your request body or limit you to specific endpoints. Any endpoint the provider supports will work:
- Chat completions
- Embeddings
- Image generation
- Audio (speech, transcription)
- Model listing
- Fine-tuning
- Moderations
Query parameters are also forwarded as-is.
Base URLs for SDKs
When configuring an SDK, set the base URL to include the provider prefix:
| Provider | SDK Base URL |
|---|---|
| OpenAI | https://proxy.aispendops.com/v1/openai/v1 |
| Anthropic | https://proxy.aispendops.com/v1/anthropic |
https://proxy.aispendops.com/v1/google-ai-studio |
See Providers for the full list.
Working examples
Runnable projects using these base URLs, in Python, TypeScript and .NET 8, are in the open-source examples repo: github.com/AI-SpendOps/examples.