Other Compatible Agent Tools
Connect agent clients that support a custom OpenAI-compatible or Anthropic-compatible provider.
Many agent products can use AnyInt without a dedicated preset when they support a custom compatible provider.
Check whether the tool can connect
The tool must let you configure all three values:
- a custom API base URL
- a custom API key
- a custom model ID
If the tool only supports a fixed provider list and does not expose a custom endpoint, it cannot connect directly through this compatibility path.
Choose the API family
| Tool requirement | AnyInt base URL | Authentication | Status |
|---|---|---|---|
| OpenAI Chat Completions | https://api.anyint.ai/openai/v1 | Bearer API key | Stable |
| OpenAI Responses | https://api.anyint.ai/openai/v1 | Bearer API key | Beta |
| Anthropic Messages | https://api.anyint.ai/anthropic or the versioned URL required by the client | x-api-key or the client's Anthropic token field | Stable |
| Gemini native | https://api.anyint.ai/gemini/v1beta | Bearer or the client's Gemini key field | Stable |
Prefer OpenAI Chat Completions for a generic custom-provider client unless the tool explicitly requires Anthropic Messages, OpenAI Responses, or Gemini-native behavior.
Generic OpenAI-compatible values
Provider name: AnyInt
API type: OpenAI-compatible
Base URL: https://api.anyint.ai/openai/v1
API key: your AnyInt API key
Model: a model ID visible to your keyIf the tool asks for the full chat endpoint, use:
https://api.anyint.ai/openai/v1/chat/completionsOtherwise, keep the base URL at /openai/v1 and let the client append the route.
Verify before production use
Confirm that the tool can:
- send a minimal text request
- complete a streaming response when streaming is enabled
- call tools if your workflow depends on tool calling
- display authentication, invalid-model, and rate-limit errors without hiding them
Compatibility with a standard API shape does not guarantee that every client feature works with every client version.