OpenAI releases GPT-6.1 Sol: near-Astra intelligence at one-fifth of Astra's API price
The short version
OpenAI released GPT-6.1 Sol, saying it delivers near-GPT-6-Astra intelligence for coding, computer use and professional work at one-fifth of Astra's standard API input and output token prices. GitHub Copilot, Amazon Bedrock and Codex CLI added it the same day.
What changed
The GPT-6 family gains a 6.1 Sol tier; the changelog lists the model identifier gpt-6.1-sol. GitHub has opened it to Copilot Pro+, Max, Business and Enterprise, AWS made it generally available on Bedrock, and Codex CLI 0.159.1 made it the default model.
What it means for you
If you are choosing between GPT-6 Astra and cheaper models, both OpenAI and AWS position this one as near-Astra at about a fifth of the cost. Those are vendor evaluations, so compare cost per task on your own workload before switching.
How to try it: check the GitHub Copilot model picker (the rollout is gradual), or choose gpt-6.1-sol in the Amazon Bedrock console or in Codex. Availability inside ChatGPT depends on plan and workspace settings.
Read with care: our automated fetch of OpenAI's own web page was blocked, so the first-party evidence here is OpenAI's news feed, its changelog and the Codex release notes. The press-reported error-rate figures and plan list are marked unconfirmed; treat OpenAI's post as authoritative.
Fact check
- VerifiedOpenAI describes GPT-6.1 Sol as near-Astra intelligence for coding, computer use and professional work, at one-fifth of Astra's standard API input and output token pricesBased on the title and summary of the post in OpenAI's news feed. OpenAI's web page returned 403 to our automated fetch and the editor could not open the full text, so benchmark figures inside it were not checked individually.Evidence
- VerifiedOpenAI's feed lists the post at 10:00 GMT on 2026-09-29 (18:00 Beijing time)Evidence
- VerifiedThe ChatGPT and Codex changelog entry for 2026-09-29 lists a new model, gpt-6.1-sol, described as near-Astra performance at lower cost than Astra, with availability depending on plan and workspace settingsEvidence
- VerifiedCodex CLI 0.159.1 made GPT-6.1 Sol the default model in the bundled catalog and in the Amazon Bedrock Mantle and Runtime catalogsEvidence
- VerifiedGitHub says GPT-6.1 Sol is generally available and rolling out gradually in GitHub Copilot for Copilot Pro+, Max, Business and Enterprise users, billed at provider list pricingGitHub notes that Business and Enterprise administrators control access through model policy; not seeing it yet is normal during the gradual rollout.Evidence
- VerifiedAWS says GPT-6.1 Sol is generally available on Amazon Bedrock and that it matches GPT-6 Astra on DeepSWE v1.1 at roughly one-fifth of the cost per taskThis is vendor benchmarking relayed by AWS, not independent testing.Evidence
- UnconfirmedTechCrunch reports the model is available immediately to Plus, Pro, Business, Enterprise and Edu users in ChatGPT Work and Codex, and not yet in ChatReported by the press only; we could not open OpenAI's post to confirm the plan list. Rely on the changelog's wording that availability depends on plan and workspace settings.Evidence
- UnconfirmedTechCrunch says the factual error rate at low reasoning effort drops from 11.4% to 7.7%, and stays within 1.9% of GPT-6 Astra across reasoning settingsFigures come from press coverage and could not be checked on OpenAI's own page; consult OpenAI's post before citing them.Evidence
Coverage timeline
- Primary sourceOpenAIIntroducing GPT-6.1 Sol
- Primary sourceOpenAI (feed)OpenAI News feed entry: Introducing GPT-6.1 Sol
- CoverageGitHubGPT-6.1 Sol in GitHub Copilot
- CoverageTechCrunchOpenAI launches GPT-6.1 Sol, says it nearly matches GPT-6 Astra and costs less
- CoverageAWSBring near-Astra intelligence to everyday work with GPT-6.1 Sol on Amazon Bedrock
- Primary sourceOpenAI (GitHub)Codex CLI 0.159.1
- Primary sourceOpenAI (changelog)ChatGPT & Codex changelog