Quick answer
- GPT-6.1 Sol (recommended)
- $2 in / $10 out per MTok
- GPT-6 Astra
- $10 in / $50 out per MTok
- GPT-6 Luna
- $0.10 in / $0.50 out per MTok
- Verified
- September 30, 2026
Your AI workspace for shipping software.
Download AIDEN free and point it at your existing Claude Code or Codex setup. No credit card, running in minutes.
Download AIDEN freeFree to start · macOS 12+ · No credit card required
Current GPT-6 roster
| Model | API identifier | Price | Context | Reasoning |
|---|---|---|---|---|
| GPT-6.1 Sol | gpt-6.1-sol | $2 in / $10 out per MTok | 1,050,000 tokens (128K max output) | low to max |
| GPT-6 Astra | gpt-6-astra | $10 in / $50 out per MTok | 1,050,000 tokens (128K max output) | low to max |
| GPT-6 Luna | gpt-6-luna | $0.10 in / $0.50 out per MTok | 1,050,000 tokens (128K max output) | none to max |
Availability and caveats
OpenAI's latest-model documentation presents Astra, Sol, and Luna as the GPT-6 family. Since September 29, 2026, its recommended Codex model for complex coding is GPT-6.1 Sol, the newer Sol: near-Astra performance at $2 in / $10 out per MTok, against $10 in / $50 out per MTok for Astra. The rollout covers Plus, Pro, Business, Enterprise and Edu; Enterprise and Edu admins must enable it, and Free and Go plans are not included at launch (OpenAI points them to GPT-6 Luna). In Codex, GPT-6.1 Sol also has an Ultrafast mode on Pro $500 and eligible Enterprise and Edu plans. Regional product pickers and staged rollouts can differ, so this page does not claim that every account sees every model on the same day.
| Older model | Status in Codex with ChatGPT sign-in |
|---|---|
| GPT-6 Sol | The previous Sol, still selectable ($2 in / $10 out per MTok) |
| GPT-5.6 Sol, Terra and Luna | Remain available during the GPT-6 rollout |
| GPT-5.5 | Retires October 14, 2026; the OpenAI API keeps it |
| GPT-5.4 and GPT-5.4 mini | Retired August 31, 2026 |
OpenAI's named replacements and the full retirement list are in the best Codex model guide.
Official source
OpenAI latest-model guide, the Codex models page for availability and retirements, plus the linked official model pages for each row.