> For the complete documentation index, see [llms.txt](https://docs.zerve.ai/zerve-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.zerve.ai/zerve-docs/agents-in-the-workspace/supported-models.md). # Supported models The model providers and models Zerve's agents can use. Zerve's agents can use models from five providers, or your own LLM gateway. Your organization's admins choose which providers and models are available. {% hint style="info" %} This list was last updated on 2 October 2026. New models are added regularly; the list your admins see in **Organization Settings → Model Settings** is always the most current. {% endhint %} ## Providers | Provider | Notes | | ------------------------ | --------------------------------------------------------------------------------------------------------------- | | **Anthropic** | Claude models. The default provider. | | **OpenAI** | GPT models | | **Google Gemini** | Gemini models | | **Amazon Bedrock** | Claude models through your own AWS account. Available when your organization's compute runs on AWS credentials. | | **OpenRouter** | Access to 300+ models from many vendors | | **Your own LLM gateway** | Any OpenAI-compatible gateway, for organizations that route all model traffic through their own endpoint | When more than one provider is approved, Zerve chooses the default provider in this order: Amazon Bedrock, Anthropic, OpenAI, Google Gemini, then OpenRouter. ## Models The **Default effort** column shows each model's default [reasoning effort](/zerve-docs/agents-in-the-workspace/models-and-effort.md#reasoning-effort), which you can change. ### Anthropic | Model | Default effort | | ----------------- | -------------- | | Claude Fable 5.1 | High | | Claude Fable 5 | High | | Claude Opus 5.5 | High | | Claude Opus 5 | High | | Claude Opus 4.8 | High | | Claude Opus 4.7 | High | | Claude Opus 4.6 | High | | Claude Opus 4.5 | High | | Claude Sonnet 5.5 | Medium | | Claude Sonnet 5 | Medium | | Claude Sonnet 4.6 | Medium | | Claude Sonnet 4.5 | Medium | | Claude Haiku 4.5 | Low | ### OpenAI | Model | Default effort | | ------------- | -------------- | | gpt-6-astra | High | | gpt-5.6-sol | Medium | | gpt-5.6-terra | Medium | | gpt-5.6-luna | Medium | | gpt-5.5 | Medium | | gpt-5.2-pro | Medium | | gpt-5.2 | Medium | | gpt-5 | Medium | | gpt-5-mini | Low | | gpt-5-nano | Low | | gpt-4.1 | Medium | | gpt-4.1-mini | Low | | gpt-4.1-nano | Low | | gpt-4o | Medium | | gpt-4o-mini | Low | ### Google Gemini | Model | Default effort | | ------------------------ | -------------- | | Gemini 3.8 Flash | Low | | Gemini 3.7 Flash | Low | | Gemini 3.6 Flash | Low | | Gemini 3.5 Flash | Low | | Gemini 3.5 Flash Lite | Low | | Gemini 3.1 Flash Lite | Low | | Gemini Flash Latest | Low | | Gemini Flash-Lite Latest | Low | | Gemini Pro Latest | Low | The "Latest" models always use Google's newest model in that family. ### Amazon Bedrock | Model | Default effort | | ----------------- | -------------- | | Claude Opus 5.5 | High | | Claude Opus 5 | High | | Claude Opus 4.8 | High | | Claude Opus 4.7 | High | | Claude Opus 4.6 | High | | Claude Sonnet 5.5 | Medium | | Claude Sonnet 5 | Medium | | Claude Sonnet 4.6 | Medium | | Claude Haiku 4.5 | Low | ### OpenRouter OpenRouter gives access to more than 300 models from dozens of vendors, including OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek, Qwen, NVIDIA, Amazon and Cohere. It also offers "Latest" models, such as Claude Sonnet Latest, that always use a vendor's newest model in a family. See [OpenRouter's model list](https://openrouter.ai/models) for what's available. ## How the model list is built Each provider's models come from that provider's own catalog. Zerve filters the list to models that work well with its agents, which need to call tools and return structured results reliably. Legacy and retired models are left out, along with some preview and open-weight models. Your admins then choose which of these models to **enable** for your organization. Users can only choose enabled models. See [Model settings](/zerve-docs/administration/model-settings.md). New organizations have **Claude Sonnet 4.6** and **Claude Haiku 4.5** enabled. ## Default models When you haven't chosen a model, Zerve uses a default for each kind of work. With Anthropic: | Kind of work | Default model | | -------------------------------- | ----------------- | | Coding, in Build and Evolve Mode | Claude Sonnet 4.6 | | Planning, in Plan mode | Claude Sonnet 4.6 | | Chat | Claude Haiku 4.5 | | Suggestions | Claude Haiku 4.5 | With OpenAI, superagents use **gpt-5.5** by default. ## Models on the Free plan On the Free plan, some models aren't available, including the largest Claude Opus and Fable models and the newest GPT models. Models available on Free include **Claude Haiku 4.5**, **Claude Sonnet 4.6** and **gpt-5.5**. See [Plans, credits and billing](/zerve-docs/administration/plans-credits-and-billing.md). ## Provider notes **Amazon Bedrock.** Bedrock uses your organization's AWS credentials rather than an API key, and models are invoked in the region configured for your account. Models that need a cross-region inference profile are handled automatically. Some newer models require provider data sharing to be enabled in your AWS account; if so, Zerve shows an error explaining what to change. **Research.** Evolve Mode's [research specialist](/zerve-docs/agents-in-the-workspace/evolve-mode/specialists.md#research) is only available with Anthropic or Amazon Bedrock. **Model quirks.** Some models have limits, such as lower output limits or not supporting certain features together. Zerve works around these automatically, so you don't need to do anything. ## Costs Models differ widely in cost. See [How credits are used](/zerve-docs/administration/how-credits-are-used.md) for prices and how they turn into credits. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.zerve.ai/zerve-docs/agents-in-the-workspace/supported-models.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.