Choosing models
You can assign a different model to each surface. All of it is configured in one place — Settings → General → AI models — and getting it right is the difference between an answer that lands and one that arrives too late to use.
Where models are set
The AI models card in Settings → General holds two layers of choice. Quick model and Reasoning model are the pair behind the control bar's tier toggle, and Models per support type assigns an OpenAI model to each surface — Chat, Coding, and System design. There is no model picker in the interview setup form.
Your per-surface choices are applied when an interview starts, and each surface keeps its model for the whole session. The Quick/Reasoning toggle on the control bar affects only surfaces you leave on the catalog default — a surface with an explicit model ignores it. To run a Claude model, leave the surface on the default and pick Claude as your Quick or Reasoning model; the per-surface dropdowns themselves are OpenAI-only.
Available models
| Model | Provider | Best for |
|---|---|---|
| GPT-5.6 Sol | OpenAI | Frontier-depth reasoning for the hardest rounds |
| GPT-5.6 Terra | OpenAI | Balanced depth and latency for everyday rounds |
| GPT-5.6 Luna | OpenAI | High-volume speed at the lightest cost |
| GPT-5.5 | OpenAI | Deep reasoning across long sessions |
| GPT-5.5 Pro | OpenAI | The deepest reasoning, when latency does not matter |
| GPT-5.4 Pro | OpenAI | Extended reasoning for architecture rounds |
| GPT-5.4 | OpenAI | System design and complex technical reasoning |
| GPT-5.4 Mini | OpenAI | Fast behavioral answers where latency is visible |
| GPT-5.4 Nano | OpenAI | The fastest answers at the lightest cost |
| GPT-5.3 Codex | OpenAI | Live coding rounds and algorithm problems |
| GPT-5.2 Pro | OpenAI | Extended reasoning on the 5.2 generation |
| GPT-5.2 | OpenAI | A balanced general-purpose option |
| GPT-5.1 | OpenAI | A stable older generation, still capable |
| GPT-4.1 Mini | OpenAI | Lowest latency and lightest credit cost |
| Claude Opus 4.6 | Anthropic | Deep architecture reasoning and nuanced tradeoffs |
| Claude Sonnet 4.6 | Anthropic | A strong all-rounder for coding and behavioral |
Gemini 2.5 Pro and Grok 3 are listed in the pickers as coming soon and cannot be selected yet. The Claude models are chosen as your Quick or Reasoning model; the per-surface dropdowns list the OpenAI models.
A pairing that works
| Surface | Model | Why |
|---|---|---|
| ⌘1 | GPT-5.4 Mini | In a behavioral round, a three-second pause reads as hesitation. Speed wins. |
| ⌘2 | GPT-5.3 Codex | Purpose-built for code; strongest on algorithm problems. |
| ⌘3 | GPT-5.4 | Design rounds reward depth, and you have time to think anyway. |
Prefer Claude for design rounds? Leave the System design surface on the catalog default and set Claude Opus 4.6 as your Reasoning model — the control bar's Reasoning tier then runs it on that surface.
The tradeoff to think about
Reasoning models produce better answers and take longer to produce them. That cost is invisible in a system design round, where you are meant to be thinking, and very visible in a rapid behavioral exchange. Match the model to the pace of the round, not to a general idea of which is "best".