Multi-Model Configuration

Approaches for configuring model-switching behavior to optimize for cost, performance, and results.

Multi-Model Configuration

goose supports several approaches for using different models within a single session, allowing you to optimize for cost, performance, and task specialization. Strategies range from manual or turn-based model selection to dynamic, context-aware switching.

📚 Documentation & Guides

- [Switching Models Mid-Session](/agents/goose/guides/goose-cli-commands/#slash-commands)：Use the /model command to change models without losing your session, keeping the same provider or switching to another. - [Subagents](/agents/goose/guides/context-engineering/subagents/)：Delegate focused tasks to isolated goose instances, each able to run its own provider and model. - [Context Engineering](/agents/goose/guides/context-engineering/)：Shape what goose keeps in context across a session, including compaction and persistent instructions.

📝 Featured Blog Posts

- [Treating LLMs Like Tools in a Toolbox: A Multi-Model Approach to Smarter AI Agents](https://goose-docs.ai/blog/2025/06/16/multi-model-in-goose)：LLMs are specialized tools, and multi-model approaches create smarter, more efficient AI agents. - [The AI Skeptic's Guide to Context Windows](https://goose-docs.ai/blog/2025/08/18/understanding-context-windows)：Learn practical ways to manage context windows and token usage in long-running sessions.

🎥 More Videos