Catalog / Categories

Model selection

Which model to pick for a job, and what it costs.

Entries

• Claude Opus 5.5

  Anthropic's recommended default model for long-running coding, agent work and document-heavy knowledge work, with a 1M-token context window.

  Model by Anthropic Claude apps, Claude Code, Any agent

  Official Reviewed — safe to install Guidance only

• DeepSeek-V4-Pro

  DeepSeek's open-weight flagship under the MIT license, for teams that need to self-host or want a low-cost hosted API for agent and coding work.

  Model by DeepSeek Any agent

  Official Reviewed — safe to install Guidance only

• Gemini 3.8 Flash

  Google's newest stable Gemini model, built for agents and long software tasks at Flash pricing, with 1M-token multimodal input.

  Model by Google Any agent

  Official Reviewed — safe to install Guidance only

• GPT-6 Astra

  OpenAI's most capable API model, recommended for complex reasoning, coding, computer use, research and document creation.

  Model by OpenAI Any agent

  Official Reviewed — safe to install Guidance only

Curated by Andrew Borden-Chisholm. Notes on building with these tools are on the blog.