> ## 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.
# Supported providers and gateways
> Which LLM providers and gateways the Stigg Agent SDK wraps, and what it meters on each.
} href="/documentation/ai-agents/llm-wrappers/providers/anthropic">
Messages API, streaming included, with the full prompt-caching breakdown
} href="/documentation/ai-agents/llm-wrappers/providers/openai">
Chat Completions and Responses API, streaming included
} href="/documentation/ai-agents/llm-wrappers/providers/gemini">
One-shot calls and chats, streaming included, with thinking tokens reported separately
} href="/documentation/ai-agents/llm-wrappers/providers/openrouter">
Gateway: bills the credits actually charged and the model that actually served, through the OpenRouter or OpenAI client
L} href="/documentation/ai-agents/llm-wrappers/providers/litellm">
Self-hosted or cloud gateway: bills the cost LiteLLM computed for each call
} href="/documentation/ai-agents/llm-wrappers/providers/groq">
Groq's own client or the OpenAI client, reported under Groq's own event name
} href="/documentation/ai-agents/llm-wrappers/providers/mistral">
Chat, fill-in-the-middle and agent calls, streaming included
} href="/documentation/ai-agents/llm-wrappers/providers/ollama">
Models you host yourself: meter and limit your own capacity
B} href="/documentation/ai-agents/llm-wrappers/providers/bedrock">
Gateway: the Converse API, for every model Bedrock serves
T} href="/documentation/ai-agents/llm-wrappers/providers/typesafe">
Typed question answering through the TypeSafe client
**Also works with any OpenAI-compatible endpoint**, through the OpenAI client:
[How OpenAI-compatible support works](/documentation/ai-agents/llm-wrappers/providers/openai-compatible)
`stigg.wrap(client)` recognizes each of these from the client object itself, so you do not name the vendor. Each provider's page shows how to wrap it and exactly what is reported.
## Why some vendors use the OpenAI client
Many vendors and gateways, including OpenRouter, LiteLLM Proxy, Groq, Mistral and Ollama, expose an API in **OpenAI's format**, and support calling it with OpenAI's `openai` client, alongside or instead of a client of their own. So you wrap an `OpenAI` client whose `baseURL` points at that vendor. Here `OpenAI` names the format of the API, not who you are calling or paying.
The SDK then picks the recipe from the address:
| The client points at | Recipe | Events | Reports |
| - | - | - | - |
| `openrouter.ai` | OpenRouter, selected automatically | `openrouter.*` | Tokens, `cost`, and the model that served |
| Your LiteLLM Proxy | LiteLLM, passed with your proxy's URL | `litellm.*` | Tokens, `cost`, and the model that served |
| `api.groq.com` | Groq, selected automatically | `groq.*` | Tokens |
| Anything else in OpenAI's format, including Mistral and Ollama | OpenAI | `openai.*` | Tokens |
Where a vendor has its own client, such as `@openrouter/sdk`, `groq-sdk`, `@mistralai/mistralai` or `ollama`, you can wrap that client instead. The SDK recognizes it by its shape.
## At a glance
| Provider | Event name | Metrics reported |
| - | - | - |
| [Anthropic](/documentation/ai-agents/llm-wrappers/providers/anthropic) | `anthropic.messages.create` | `input_tokens`, `output_tokens`, `cache_read_input_tokens`, `cache_creation_input_tokens`, `total_tokens` |
| [OpenAI](/documentation/ai-agents/llm-wrappers/providers/openai), Chat Completions | `openai.chat.completions.create` | `prompt_tokens`, `completion_tokens`, `cached_tokens`, `total_tokens` |
| [OpenAI](/documentation/ai-agents/llm-wrappers/providers/openai), Responses | `openai.responses.create` | `input_tokens`, `output_tokens`, `cached_tokens`, `total_tokens` |
| [Google Gemini](/documentation/ai-agents/llm-wrappers/providers/gemini) | `gemini.models.generateContent`, `gemini.models.generateContentStream` | `promptTokenCount`, `candidatesTokenCount`, `cachedContentTokenCount`, `thoughtsTokenCount`, `toolUsePromptTokenCount`, `totalTokenCount` |
| [Google Gemini](/documentation/ai-agents/llm-wrappers/providers/gemini), chats | `gemini.chats.sendMessage`, `gemini.chats.sendMessageStream` | The same as Gemini above |
| [Groq](/documentation/ai-agents/llm-wrappers/providers/groq) | `groq.chat.completions.create` | `prompt_tokens`, `completion_tokens`, `cached_tokens`, `total_tokens` |
| [Mistral](/documentation/ai-agents/llm-wrappers/providers/mistral) | `mistral.chat.complete`, `mistral.chat.stream`, `mistral.chat.parse`, `mistral.chat.parseStream`, `mistral.fim.complete`, `mistral.fim.stream`, `mistral.agents.complete`, `mistral.agents.stream` | `prompt_tokens`, `completion_tokens`, `total_tokens` |
| [Ollama](/documentation/ai-agents/llm-wrappers/providers/ollama) | `ollama.chat`, `ollama.generate` | `prompt_eval_count`, `eval_count`, `total_tokens` |
| [OpenRouter](/documentation/ai-agents/llm-wrappers/providers/openrouter), Chat Completions | `openrouter.chat.send`, `openrouter.chat.completions.create` | `prompt_tokens`, `completion_tokens`, `cached_tokens`, `total_tokens`, `cost` |
| [OpenRouter](/documentation/ai-agents/llm-wrappers/providers/openrouter), Responses | `openrouter.responses.send`, `openrouter.responses.create` | `input_tokens`, `output_tokens`, `cached_tokens`, `total_tokens`, `cost` |
| [LiteLLM Proxy](/documentation/ai-agents/llm-wrappers/providers/litellm), Chat Completions | `litellm.chat.completions.create` | `prompt_tokens`, `completion_tokens`, `cached_tokens`, `total_tokens`, `cost` (non-streaming calls only) |
| [LiteLLM Proxy](/documentation/ai-agents/llm-wrappers/providers/litellm), Responses | `litellm.responses.create` | `input_tokens`, `output_tokens`, `cached_tokens`, `total_tokens`, `cost` (non-streaming calls only) |
| [Amazon Bedrock](/documentation/ai-agents/llm-wrappers/providers/bedrock) | `bedrock.send` | `inputTokens`, `outputTokens`, `cacheReadInputTokens`, `cacheWriteInputTokens`, `totalTokens` |
| [TypeSafe](/documentation/ai-agents/llm-wrappers/providers/typesafe) | `typesafe.systemOne` | `input_tokens`, `output_tokens`, `total_tokens` |
| [OpenAI-compatible](/documentation/ai-agents/llm-wrappers/providers/openai-compatible) | `openai.chat.completions.create`, `openai.responses.create` | The same as OpenAI |
For any other vendor, including your own models, see [Bring your own vendor](/documentation/ai-agents/llm-wrappers/providers/custom).
## How usage is reported
* **Same names as the vendor.** Each metric keeps the name the vendor uses, such as `input_tokens` for Anthropic, `prompt_tokens` for OpenAI or `promptTokenCount` for Gemini. You can compare the numbers in Stigg directly with the vendor's own usage reports, such as its usage dashboard.
* **A total, and its parts.** Every vendor also reports a total: the vendor's own where it publishes one, such as `total_tokens` or `totalTokenCount`, and otherwise the sum of the parts, added up by the SDK (Anthropic, Ollama and TypeSafe). The total includes the other counters, so price the total or the parts, never both.
* **Every usage field, not just the named metrics.** Each event also includes every usage field the vendor returned, with a `raw.` prefix (`raw.`). If a vendor adds a field the SDK does not report as a metric yet, you can build a meter on it without waiting for an SDK update.
* **No prompts or responses.** Only usage numbers are sent to Stigg. The text of your prompts and the model's responses never leaves your application.
## Related topics
- [Gateways](/documentation/ai-agents/llm-wrappers/gateways.md)
- [LLM and gateway wrappers](/documentation/ai-agents/llm-wrappers/overview.md)
- [OpenAI](/documentation/ai-agents/llm-wrappers/providers/openai.md)
- [LiteLLM Proxy](/documentation/ai-agents/llm-wrappers/providers/litellm.md)
- [OpenAI-compatible endpoints](/documentation/ai-agents/llm-wrappers/providers/openai-compatible.md)
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.