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

# Supported AI Models and Providers Available in Portrix

> Browse all 400+ AI models available through Portrix, organized by provider. Includes model IDs, context windows, and supported capabilities.

Portrix provides access to 400+ models from leading AI providers through a single, unified API. All models are accessed using the `{provider}/{model}` naming format — for example, `openai/gpt-4o` or `anthropic/claude-3-5-sonnet`.

<Note>
  The full list of available models is always up to date via the API: `GET /v1/models`. This page shows a curated overview of the most popular models. New models are added continuously as providers release them.
</Note>

## How to list models programmatically

Fetch the live model catalog at any time using the `/v1/models` endpoint:

<CodeGroup>
  ```bash cURL theme={null}
  curl https://api.portrix.ai/v1/models \
    -H "Authorization: Bearer YOUR_PORTRIX_KEY"
  ```

  ```python Python theme={null}
  import openai

  client = openai.OpenAI(
      api_key="YOUR_PORTRIX_KEY",
      base_url="https://api.portrix.ai/v1",
  )

  models = client.models.list()
  for model in models.data:
      print(model.id)
  ```
</CodeGroup>

***

## OpenAI

| Model ID                        | Context Window | Capabilities        | Notes                                   |
| ------------------------------- | -------------: | ------------------- | --------------------------------------- |
| `openai/gpt-4o`                 |           128k | Chat, Vision, Tools | Flagship multimodal model               |
| `openai/gpt-4o-mini`            |           128k | Chat, Tools         | Cost-effective; ideal for high-volume   |
| `openai/o3-mini`                |           128k | Chat, Reasoning     | Extended reasoning and complex problems |
| `openai/gpt-4-turbo`            |           128k | Chat, Vision, Tools | Previous flagship; strong performance   |
| `openai/text-embedding-3-small` |             8k | Embeddings          | Best balance of speed and quality       |
| `openai/text-embedding-3-large` |             8k | Embeddings          | Highest quality embeddings from OpenAI  |

***

## Anthropic

| Model ID                      | Context Window | Capabilities        | Notes                                        |
| ----------------------------- | -------------: | ------------------- | -------------------------------------------- |
| `anthropic/claude-3-5-sonnet` |           200k | Chat, Vision, Tools | Top-tier reasoning and instruction following |
| `anthropic/claude-3-5-haiku`  |           200k | Chat, Vision, Tools | Fast and cost-effective; great for agents    |
| `anthropic/claude-3-opus`     |           200k | Chat, Vision, Tools | Maximum capability for complex tasks         |

***

## Google

| Model ID                  | Context Window | Capabilities        | Notes                                        |
| ------------------------- | -------------: | ------------------- | -------------------------------------------- |
| `google/gemini-2.0-flash` |             1M | Chat, Vision, Tools | Very fast; excellent multimodal performance  |
| `google/gemini-1.5-pro`   |             1M | Chat, Vision, Tools | Best for long-context document understanding |
| `google/gemini-1.5-flash` |             1M | Chat, Vision, Tools | Optimized for speed at 1M token context      |

***

## Mistral

| Model ID                | Context Window | Capabilities | Notes                                  |
| ----------------------- | -------------: | ------------ | -------------------------------------- |
| `mistral/mistral-large` |            32k | Chat, Tools  | Mistral's most powerful model          |
| `mistral/mistral-small` |            32k | Chat, Tools  | Fast and affordable for everyday tasks |
| `mistral/codestral`     |            32k | Chat, Tools  | Specialized for code generation        |

***

## Meta

| Model ID             | Context Window | Capabilities | Notes                                        |
| -------------------- | -------------: | ------------ | -------------------------------------------- |
| `meta/llama-3.3-70b` |           128k | Chat, Tools  | Latest Llama 3.3; strong open-weight model   |
| `meta/llama-3.1-8b`  |           128k | Chat         | Lightweight; fast inference for simple tasks |

***

## Cohere

| Model ID                | Context Window | Capabilities | Notes                                      |
| ----------------------- | -------------: | ------------ | ------------------------------------------ |
| `cohere/command-r-plus` |           128k | Chat, Tools  | Cohere's most capable model; great for RAG |
| `cohere/command-r`      |           128k | Chat, Tools  | Balanced performance and cost              |
| `cohere/embed-v3`       |     512 tokens | Embeddings   | High-quality multilingual embeddings       |

***

## Model capabilities key

| Capability     | Description                                                                                     |
| -------------- | ----------------------------------------------------------------------------------------------- |
| **Chat**       | Standard text generation via the `/v1/chat/completions` endpoint                                |
| **Vision**     | Accepts image inputs alongside text; supports image URLs and base64-encoded images              |
| **Tools**      | Supports function calling (tool use) for structured outputs and agentic workflows               |
| **Streaming**  | Supports server-sent events (SSE) streaming via `"stream": true` — available on all chat models |
| **Embeddings** | Converts text to vector representations via the `/v1/embeddings` endpoint                       |
| **Reasoning**  | Extended chain-of-thought reasoning for multi-step problem solving (OpenAI `o` series)          |

<Tip>
  Use model IDs exactly as shown — for example, `openai/gpt-4o` — when making API calls. Omitting the provider prefix will result in a `404 Model Not Found` error.
</Tip>