EnhancementWeb App

Chat on Claude Opus 5.5 and GPT-6 in every model tier

Deep Intelligence, Expert and Fast now answer with the latest Anthropic and OpenAI models, so your hardest questions and your quickest ones both get a newer generation behind them.

You pick a model tier in Chief based on the job: Deep Intelligence for long analysis, Expert for everyday work, Fast for quick answers. Those tiers ran on Claude Opus 5 and the GPT-5.4 and GPT-5.5 family, and they now move to the newest generation.

Deep Intelligence on Opus 5.5 and GPT-6 Astra

Deep Intelligence now runs on Claude Opus 5.5 when you choose Anthropic and GPT-6 Astra when you choose OpenAI. Agents use these models for deep reasoning and full analysis on longer-running tasks.

Expert and Fast on GPT-6

With OpenAI selected, Expert moves from GPT-5.4 Mini to GPT-6 Sol, and Fast moves from GPT-5.4 Nano to GPT-6 Luna. Anthropic's Sonnet 5 and Haiku 4.5, and Google's Gemini models, stay as they were.

Updated vendor selector

The vendor selector lists the new names next to each tier, so you can see which model answers you:

  • Anthropic: 4.5 Haiku (Fast), Sonnet 5 (Expert), Opus 5.5 (Deep Intelligence)
  • OpenAI: GPT-6 Luna (Fast), GPT-6 Sol (Expert), GPT-6 Astra (Deep Intelligence)

Your own Anthropic key

If you bring your own Anthropic key, Deep Intelligence conversations on Opus 5.5 keep working across multiple turns, including on newer Anthropic accounts.

Open a chat and pick your tier as usual. Chief routes it to the new models with no setup on your part.