Select from hundreds of AI models for your team and individual personas.

Choosing AI Models

Workspace supports hundreds of AI models — including Claude, GPT, Gemini, Llama, Mistral, and many more.

Default Model

Every team has a default model set in the editor's Settings panel. This model is used by all personas unless they have an override. A good default for general use is google/gemini-2.5-flash-lite — fast and cost-effective.

Per-Persona Override

Each persona block has an optional Model Override field. Use this when different agents need different capabilities:

Use CaseSuggested Model
Deep analysis, codingClaude (Sonnet/Opus)
General tasksGPT-4o
Fast responsesGemini Flash, Llama
Cost-sensitive tasksGemini Flash Lite, small models

Cost Implications

Different models have different per-token costs. More capable models cost more. You can see the cost breakdown in Billing > Usage after running your team.

Start with a cheaper model and upgrade specific personas only if you need better quality for their tasks. This keeps costs down while maintaining quality where it matters.

Model Names

Models are referenced by their identifier, for example:

  • anthropic/claude-sonnet-4
  • openai/gpt-4o
  • google/gemini-2.5-flash-lite
  • meta-llama/llama-3.1-70b-instruct

The model picker in the editor shows available models with their names and providers.

Adding PersonasTeam Settings