OpenAI · GPT-5.6 · 9 July 2026
The model name changed. The practical choice is still effort.
GPT-5.5 Instant remains the everyday default, while GPT-5.6 Sol now powers Medium, High and Extra High reasoning and Sol Pro powers Pro.
Learn the effort ladder, not the release branding. We will test whether higher effort improves the jobs our audience actually does.
Why this may matter
Most people do not need to memorise four celestial model names. They need to match effort to consequence: Instant for reversible work, more reasoning for difficult analysis, and Pro only when the additional time and allowance can earn their keep.
Extra reasoning can make an answer more elaborate without making its evidence true. Verify the source and decision, not the confidence of the prose.
Inspect the change, example and test detail
What changed
OpenAI began rolling out GPT-5.6 Sol and Sol Pro in ChatGPT. Sol, Terra and Luna are available in Work and Codex on eligible plans, while the API offers Sol, Terra and Luna; Terra and Luna are not selectable in standard ChatGPT conversations.
Real example
Draft a meeting summary with Instant. Use Medium or High to compare conflicting contracts. Reserve Pro for a difficult, long-running workflow where a weaker attempt has already failed.
A bounded test
Take one real decision and run it at Instant, Medium and High with the same brief. Record useful differences, elapsed time and how much you corrected.
Sources and review
Sources retain their own provenance. A provider announcement establishes what was announced; it does not become independent testing.
Canonical claim: OpenAI began rolling out GPT-5.6 Sol and Sol Pro in ChatGPT. Sol, Terra and Luna are available in Work and Codex on eligible plans, while the API offers Sol, Terra and Luna; Terra and Luna are not selectable in standard ChatGPT conversations.
Readiness: Use the dated guidance with the stated human check; this is not a completed Human Bit test.
- Released
- 9 July 2026
- Reviewed
- 25 July 2026
- Review by
- 8 August 2026
- Review owner
- The Human Bit editorial
- Revision
- gpt-56-reasoning-family@v1
- Volatility
- volatile