OpenAI's mid-tier GPT-6-family model, sitting below GPT-6 Astra — replacing GPT-6 Sol at the same input/output price with a cheaper cache-read rate.
GPT-6.1 Sol bills $2 / 1M input and $10 / 1M output for prompts up to 272K tokens; longer prompts are billed at $4 / $15 per 1M. Cached input reads are $0.10 / 1M (≤272K) or $0.20 / 1M (>272K) — half GPT-6 Sol's cache price at the same base rate.
OpenAI released GPT-6.1 Sol on September 29, 2026 at DevDay, alongside GPT-6 Luna and the new $500/mo ChatGPT Pro 500 tier. It extends the GPT-6 family below GPT-6 Astra, giving developers and ChatGPT Work/Codex users on Plus, Pro, Business, Enterprise, and Edu plans a cheaper option in the new generation without stepping down to the GPT-5.6 family.
GPT-6.1 Sol replaces GPT-6 Sol at the identical $2/$10 (and $4/$15 long-context) price. The one change is cache pricing: cached input reads were cut 50%, from $0.20 to $0.10 per 1M tokens (≤272K context) — a meaningful saving for cache-heavy agentic workloads.
Sol sits as the balanced, general-purpose pick in the GPT-6 family: strong coding and agentic performance without Astra's premium pricing. As with prior OpenAI launches, expect model-picker availability on consumer ChatGPT plans to keep evolving during the rollout window.
| Model | Input / 1M | Output / 1M | Context |
|---|---|---|---|
| GPT-6.1 Sol | $2 | $10 | 1M |
| GPT-6 Astra | $10 | $50 | 1.05M |
| GPT-6 Luna | $0.10 | $0.50 | 1M |
| GPT-5.6 Sol | $4 | $20 | 1M |
| Claude Sonnet 5.5 | $2 | $10 | 1M |
GPT-6.1 Sol matches Claude Sonnet 5.5 exactly on price ($2/$10) and undercuts GPT-5.6 Sol at half the input cost — a sign OpenAI is pricing the new GPT-6 generation's mid-tier below its predecessor rather than at a premium.