GPT-6 Sol
GPT-6 Sol is OpenAI's September 22, 2026 GPT-6 workhorse. The API ID is gpt-6-sol, with a 1,050,000-token context window, 128,000 maximum output, text and image input, text output, and reasoning effort from none through max.
Sol brings much of the GPT-6 agent and coding stack to a lower-cost tier than Astra. The important decision is not only benchmark rank: long-prompt multipliers, cache behavior, tool access, and the actual ChatGPT Work, Codex, or API surface can change both cost and fit.
OpenAI's launch and live model page control identity, limits, pricing, modalities, tools, and availability. OpenAI's benchmark and factuality comparisons are provider-reported. Community launch threads supply questions about rollout, usage limits, and the 272K price threshold, not model facts.
- Run complex repository and coding-agent workflows.
- Automate professional tasks across tools and applications.
- Use computer interaction with explicit permissions and verification.
- Route work that needs more capability than Luna without Astra pricing.
The base rate is $2 input, $0.20 cached input, $2.50 cache write, and $10 output per million tokens. Once input exceeds 272,000 tokens, OpenAI bills 2x input and cache rates plus 1.5x output for the full request, so a 1.05M context window is capacity rather than a flat-price invitation.
OpenAI lists gpt-6-sol in the API and launched it in ChatGPT Work and Codex for Plus, Pro, Business, Enterprise, and Edu users. The launch also says the models were not yet in Chat and were rolling out gradually, so product, plan, account, and API availability remain separate checks.
Astra remains OpenAI's highest-capability GPT-6 model. Sol is the general workhorse for complex coding and professional tasks; Luna is the high-volume efficiency tier. Route by validated task success and total cost instead of assuming one family member wins every workload.
GPT-6 Sol FAQ
Common questions about GPT-6 Sol.
What is the GPT-6 Sol API model ID?+
Use gpt-6-sol. The current model page documents text and image input, text output, a 1.05M context window, and up to 128K output tokens.
How much does GPT-6 Sol cost?+
Base rates are $2 input, $0.20 cached input, $2.50 cache write, and $10 output per million tokens. Requests above 272K input tokens use higher rates for the full request.
Should I use GPT-6 Sol or Astra?+
Start with Sol when validated task quality is sufficient and cost or usage headroom matters. Use Astra when the hardest work benefits enough from its higher capability to justify the price and access boundary.