A model guide for the GPT-6 family
OpenAI's practical guide to the GPT-6 family (Astra, 6.1 Sol, Luna), with per-token pricing and when to use each model. It gives guidance on picking a reasoning level (low to max), Fast and Ultrafast speed modes, and rewriting skills, AGENTS.md and prompts with clear decision boundaries and a definition of "done". For long-running agent work it covers mid-turn steering over the Responses WebSocket API, async tool calling, multi-agent delegation (beta) and computer use with Playwright or PyAutoGUI. It also gives production advice on prompt caching, compaction, monitoring and measuring cost per successful task.
TL;DR OpenAI's how-to for choosing between GPT-6 Astra, Sol and Luna, setting reasoning effort, prompting, running long agent tasks and controlling cost in production.
Organisations named here
Technologies named here
- GPT-6 Astra
- Codex
- GPT-6.1 Sol
- Responses API
- GPT-6 Luna
- Responses WebSocket API
- Prompt caching
- Compaction
- Computer use tool
- Playwright
- PyAutoGUI
- Fast mode
- Ultrafast mode
- Caching dashboard and diagnostics guide
- AGENTS.md
- OpenAI API
Techniques named here
- Reasoning effort selection
- Mid-turn steering
- Asynchronous tool calling
- Multi-agent delegation
- Prompt caching
- compaction
- Computer use via code execution harness
Related in prompting
Dicklesworthstone/skillranker
miqdadbadjuber/anti-slop
