> ## Documentation Index
> Fetch the complete documentation index at: https://docs.stigg.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Providers and frameworks

> The providers, AI gateways and agent frameworks the Stigg Agent SDK supports, and how to add your own.

You add Stigg where your code already calls models. Wrap the client you use to call a provider or AI gateway, wrap `fetch` for a gateway you call with plain HTTP, or add Stigg to the framework that runs your agents. For how each one works, see [How it works](https://docs.stigg.io/documentation/ai-agents/overview.md#how-it-works). Each page below shows the setup steps and exactly what is reported.

## Providers

`stigg.wrap(client)` recognizes each of these from the client object itself, so you do not name the vendor.

> [card] [Anthropic](https://docs.stigg.io/documentation/ai-agents/providers/anthropic.md)
>
> Messages API, streaming included, with the full prompt-caching breakdown

> [card] [OpenAI](https://docs.stigg.io/documentation/ai-agents/providers/openai.md)
>
> Chat Completions and Responses API, streaming included

> [card] [Google Gemini](https://docs.stigg.io/documentation/ai-agents/providers/gemini.md)
>
> One-shot calls and chats, streaming included, with thinking tokens reported separately

> [card] [Mistral](https://docs.stigg.io/documentation/ai-agents/providers/mistral.md)
>
> Chat, fill-in-the-middle and agent calls, streaming included

> [card] [Groq](https://docs.stigg.io/documentation/ai-agents/providers/groq.md)
>
> Groq's own client or the OpenAI client, reported under Groq's own event name

> [card] [xAI](https://docs.stigg.io/documentation/ai-agents/providers/xai.md)
>
> Grok models, through the OpenAI client

> [card] [DeepSeek](https://docs.stigg.io/documentation/ai-agents/providers/deepseek.md)
>
> DeepSeek's own models, through the OpenAI client

> [card] [Perplexity](https://docs.stigg.io/documentation/ai-agents/providers/perplexity.md)
>
> Models with built-in web search, through the OpenAI client

> [card] [Together AI](https://docs.stigg.io/documentation/ai-agents/providers/together.md)
>
> A large catalog of open-source models, through the OpenAI client

> [card] [Fireworks AI](https://docs.stigg.io/documentation/ai-agents/providers/fireworks.md)
>
> Open-source and fine-tuned models, through the OpenAI client

> [card] [Ollama](https://docs.stigg.io/documentation/ai-agents/providers/ollama.md)
>
> Models you host yourself: meter and limit your own capacity

> [card] [vLLM](https://docs.stigg.io/documentation/ai-agents/providers/vllm.md)
>
> Your self-hosted model server, through the OpenAI client

> [card] [TypeSafe](https://docs.stigg.io/documentation/ai-agents/providers/typesafe.md)
>
> Jev, a model that answers typed questions about your data (yes or no, choice or score) in one call

## AI gateways

> [card] [OpenRouter](https://docs.stigg.io/documentation/ai-agents/providers/openrouter.md)
>
> Reports the credits actually charged and the model that actually served, through the OpenRouter or OpenAI client

> [card] [LiteLLM Proxy](https://docs.stigg.io/documentation/ai-agents/providers/litellm.md)
>
> Self-hosted or cloud gateway: reports the cost LiteLLM computed for each call

> [card] [Amazon Bedrock](https://docs.stigg.io/documentation/ai-agents/providers/bedrock.md)
>
> The Converse API, for every model Bedrock serves

> [card] [Cloudflare](https://docs.stigg.io/documentation/ai-agents/providers/cloudflare.md)
>
> AI Gateway and Workers AI, with cache hits left unbilled and the model that served

> [card] [Databricks](https://docs.stigg.io/documentation/ai-agents/providers/databricks.md)
>
> Model serving endpoints, with the model that actually answered each call

> [card] [agentgateway](https://docs.stigg.io/documentation/ai-agents/providers/agentgateway.md)
>
> The open-source, self-hosted gateway: tokens and the model that served, over a client or plain HTTP

> [card] [Solo.io](https://docs.stigg.io/documentation/ai-agents/providers/agentgateway.md)
>
> Solo.io's enterprise agentgateway and Gloo AI Gateway, through the agentgateway recipe with your gateway's URL

## Frameworks

One integration in the framework covers every agent it runs, so you do not wrap each agent's client.

> [card] [Vercel AI SDK](https://docs.stigg.io/documentation/ai-agents/orchestrators/vercel-ai-sdk.md)
>
> One middleware in `wrapLanguageModel`, for `generateText`, `streamText` and agents

> [card] [LangChain and LangGraph](https://docs.stigg.io/documentation/ai-agents/orchestrators/langchain.md)
>
> One callback handler in a run's `callbacks`, for `createAgent` graphs and hand-built `StateGraph`s

> [card] [Google ADK](https://docs.stigg.io/documentation/ai-agents/orchestrators/google-adk.md)
>
> One plugin in a Runner's `plugins`, for every agent the Runner reaches

## Bring your own

> [card] [Any OpenAI-compatible API](https://docs.stigg.io/documentation/ai-agents/providers/custom.md#any-openai-compatible-api)
>
> Point the OpenAI client at any endpoint in OpenAI's format and it is metered with no extra code

> [card] [Custom adapter](https://docs.stigg.io/documentation/ai-agents/providers/custom.md)
>
> Any other model or API, including your own: describe where its usage is and the SDK does the rest

> [card] [Your own framework](https://docs.stigg.io/documentation/ai-agents/providers/custom.md#your-own-framework)
>
> Any other framework or harness, through `stigg.middleware()` around each step


## Related topics

- [Usage metering](https://docs.stigg.io/documentation/ai-agents/usage-metering.md)
- [Bring your own vendor or framework](https://docs.stigg.io/documentation/ai-agents/providers/custom.md)
- [Vercel AI SDK](https://docs.stigg.io/documentation/ai-agents/orchestrators/vercel-ai-sdk.md)
- [Google ADK](https://docs.stigg.io/documentation/ai-agents/orchestrators/google-adk.md)
- [vLLM](https://docs.stigg.io/documentation/ai-agents/providers/vllm.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.