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

# AI models supported by WeCareRemote

> WeCareRemote supports OpenAI, Anthropic, Google, Groq, Ollama, AWS Bedrock, and more. Browse all supported models and learn how to select one for your session.

WeCareRemote gives you the flexibility to choose the AI model that best fits your use case. Rather than locking you into a single provider, the platform supports a broad range of leading models — from frontier models like GPT-4o and Claude 3.5 Sonnet to locally hosted options for privacy-sensitive environments. You can set a platform-wide default or specify a model per request.

<Tip>
  If you are unsure where to start, GPT-4o (OpenAI) and Claude 3.5 Sonnet (Anthropic) deliver strong results across a wide range of tasks including multi-language conversation, document retrieval, and nuanced role-aware responses.
</Tip>

## Supported providers

<AccordionGroup>
  <Accordion title="OpenAI" icon="openai">
    OpenAI models are widely used and well-suited for general-purpose conversation, summarization, and document Q\&A.

    | Model           | Notes                                            |
    | --------------- | ------------------------------------------------ |
    | `gpt-4o`        | Latest flagship model; fast and highly capable   |
    | `gpt-4o-mini`   | Lightweight, cost-efficient variant of GPT-4o    |
    | `gpt-4-turbo`   | High-performance model with large context window |
    | `gpt-3.5-turbo` | Fast and affordable for simpler tasks            |
  </Accordion>

  <Accordion title="Anthropic" icon="a">
    Anthropic's Claude models are known for their nuanced, thoughtful responses and strong instruction-following.

    | Model                        | Notes                                 |
    | ---------------------------- | ------------------------------------- |
    | `claude-3-5-sonnet-20241022` | Anthropic's most capable model        |
    | `claude-3-haiku-20240307`    | Fastest and most compact Claude model |
    | `claude-3-opus-20240229`     | Most powerful Claude 3 model          |
  </Accordion>

  <Accordion title="Google" icon="google">
    Google AI models include the Gemini family, available through Google AI Studio or Vertex AI.

    | Model              | Notes                                                 |
    | ------------------ | ----------------------------------------------------- |
    | `gemini-1.5-pro`   | Long context, strong reasoning and multimodal support |
    | `gemini-1.5-flash` | Fast, efficient variant of Gemini 1.5                 |
    | `gemini-pro`       | Previous-generation general-purpose model             |
  </Accordion>

  <Accordion title="Groq" icon="bolt">
    Groq provides ultra-fast inference for open-weight models, making it ideal for low-latency interactions.

    | Model                     | Notes                                       |
    | ------------------------- | ------------------------------------------- |
    | `llama-3.1-70b-versatile` | Large Llama 3.1 model via Groq              |
    | `llama-3.1-8b-instant`    | Compact Llama 3.1 for fast responses        |
    | `mixtral-8x7b-32768`      | Mixtral MoE model with large context window |
    | `gemma2-9b-it`            | Google Gemma 2 via Groq                     |
  </Accordion>

  <Accordion title="Ollama (local)" icon="house">
    Ollama lets you run open-weight models locally on your own hardware. No data leaves your environment.

    | Example models             | Notes                                                |
    | -------------------------- | ---------------------------------------------------- |
    | `llama3`                   | Meta's Llama 3                                       |
    | `mistral`                  | Mistral 7B                                           |
    | `phi3`                     | Microsoft Phi-3                                      |
    | Any Ollama-supported model | See [ollama.com/library](https://ollama.com/library) |

    <Note>
      Ollama is the recommended option for privacy-sensitive use cases. Because models run locally, conversation data never leaves your organization's infrastructure.
    </Note>
  </Accordion>

  <Accordion title="AWS Bedrock" icon="aws">
    AWS Bedrock provides managed access to a curated set of foundation models through Amazon's infrastructure.

    Models available via Bedrock include Anthropic Claude, Meta Llama, Amazon Titan, Mistral, and others depending on your region and access configuration.
  </Accordion>

  <Accordion title="Azure OpenAI" icon="microsoft">
    Access OpenAI GPT models through your organization's Azure OpenAI resource, with enterprise-grade compliance and data residency controls.

    The models available depend on the deployments configured in your Azure portal.
  </Accordion>

  <Accordion title="OpenRouter" icon="shuffle">
    OpenRouter is an aggregator that provides access to models from many providers through a single API, including models from OpenAI, Anthropic, Google, Meta, Mistral, and more.
  </Accordion>

  <Accordion title="DeepSeek" icon="magnifying-glass">
    DeepSeek offers capable models particularly well-suited for reasoning and coding tasks.
  </Accordion>

  <Accordion title="Custom OpenAI-compatible endpoints" icon="plug">
    Connect any API that is compatible with the OpenAI API format — including self-hosted models, fine-tuned endpoints, or third-party providers not listed above.
  </Accordion>
</AccordionGroup>

## How to select a model

You can choose which model the assistant uses in two ways:

**Platform default** — Your administrator configures the default model for the platform. All conversations use this model unless you override it per request. See [Configure the AI assistant](/ai-assistant/configuration) for details.

**Per request** — When interacting with the assistant through the API, pass a `model` parameter in your request body to use a specific model for that interaction only. This lets you run different models for different tasks without changing the platform default.