---
title: "Providers"
description: "The six provider types Cerevisor supports, the library that holds them, and how to mix providers within a workflow."
slug: guides/providers
section: guides
subsection: providers
canonical_url: https://cerevisor.com/docs/guides/providers
last_verified: 2026-05-18
last_verified_version: "1.2.0"
updated_at: 2026-05-18T15:08:18.053416+00:00
---

Cerevisor doesn't ship a model. You bring your own. Six provider types are supported, and you can have any number of credentials in the library.

- **[Overview](./overview.md)**: the six types, the library, defaults, resolution.
- **[Anthropic](./anthropic.md)**: API key, models, cost.
- **[Gemini](./gemini.md)**: API key from Google AI Studio, the 2.5 and 3.x families, >200K-tier pricing.
- **[Claude Agent SDK](./claude-agent-sdk.md)**: Claude Code's harness in-process, delegated-loop runs.
- **[OpenAI-compatible & local](./openai-compatible-and-local.md)**: Ollama, OpenAI, OpenRouter, Together, Groq, Fireworks, vLLM, LM Studio.
- **[OpenAI Codex CLI](./openai-codex-cli.md)**: ChatGPT subscription auth, device-auth wizard, sandbox modes.
- **[Cursor Agent](./cursor-agent.md)**: API key, local + cloud, optional dependency.
- **[Per-agent overrides](./per-agent-overrides.md)**: mixing providers within a single workflow.
- **[Managing the Provider Library](./managing-the-provider-library.md)**: add, remove, set default, deletion guards.