Guides · Models
GPT-6.1 Sol: price, context window and Codex use
| Developer | OpenAI |
|---|---|
| Released | 29 September 2026 (API, Codex and ChatGPT Work) |
| Model ID | gpt-6.1-sol |
| Context window (API) | 1,050,000 tokens (922,000 max input, 128,000 max output) |
| API price per 1M tokens | $2 input / $0.10 cached input / $10 output |
| Long prompts (over 272K input tokens) | 2x input and cache rates, 1.5x output |
| Reasoning effort (API) | low, medium (default), high, xhigh, max |
| ChatGPT | Work and Codex only, not Chat |
| Codex plans | Plus, Pro, Business, Enterprise and Edu (off by default on Enterprise and Edu) |
| Default context in Codex and Work | 500K tokens since 11 October 2026 |
GPT-6.1 Sol delivers near-Astra performance at a lower cost for complex coding, computer use, and professional work.
What OpenAI says it's for
OpenAI's model selection guide suggests GPT-6.1 Sol for complex projects where cost matters, such as building a board presentation from financial results or a website from a product brief, and recommends comparing it with Astra on the same task. Its Codex docs describe it as the pick for repeated, long-running work across code, apps and documents, with Astra kept for the most demanding work.
GPT-6.1 Sol in Codex
- Rolled out to Plus, Pro, Business, Enterprise and Edu in the desktop app, CLI and ChatGPT Work. Free and Go weren't included at launch.
- Codex CLI 0.161.0 (7 Oct 2026) made it the default in the bundled catalog. Select it with codex --model gpt-6.1-sol or model = "gpt-6.1-sol" in config.toml.
- OpenAI estimates 15 to 160 local messages per five-hour window on Plus and Standard Business. Pro currently has no five-hour limit.
- Credit rate: 50 credits per million input tokens, 2.5 per million cached input tokens and 250 per million output tokens.
- Reasoning ranges from Light to Ultra in Codex. Max and Ultra depend on your settings, and Ultra uses subagents.
- Ultrafast mode is available on Pro $500 and eligible Enterprise and Edu plans. It uses included usage at 8x the Standard rate and credits at 6x.
On 11 October, Codex lead Tibo said the default context length for GPT-6.1 Sol in Codex and Work went up to 500k tokens from 272k, to reduce compactions.
With this change, we are not charging long context fees for GPT-6.1 Sol when drawing usage from the subscription.
API notes
- Use the Responses API for tool calling. Chat Completions works without tools. The none and minimal reasoning efforts aren't supported.
- Cached input costs 5% of the input rate, cache writes 1.25x. Fast mode costs 2x Standard, Batch and Flex 50% less, and Ultrafast (service_tier "ultrafast", added 8 Oct 2026) 6x.
- US and EU data residency are supported, including with Fast and Ultrafast. Knowledge cutoff: 30 April 2026.
Independent benchmarks
Artificial Analysis lists GPT-6.1 Sol (max effort) at 52 on its Intelligence Index v4.3.2, at $0.72 per Intelligence Index task, as of 13 October 2026. That was the lowest cost per task of the models listed in its 12 October comparison, where it used about 38k output tokens per task.
Questions
- What's the difference between GPT-6.1 Sol and GPT-6 Sol?
- GPT-6.1 Sol is newer, with a later knowledge cutoff and cheaper cached input ($0.10 against $0.20 per million tokens). Input and output prices are the same, $2 / $10. GPT-6.1 Sol doesn't support the none reasoning effort; GPT-6 Sol does. OpenAI recommends GPT-6.1 Sol in Codex.
- Can I use GPT-6.1 Sol in ChatGPT's Chat tab?
- No. OpenAI says GPT-6.1 Sol is available in Work and Codex, not in Chat.
- Is GPT-6.1 Sol available on the free ChatGPT plan?
- Not at launch. Free and Go users get GPT-6 Luna in the Codex desktop app, subject to rollout.
Sources
- GPT-6.1 Sol model page · OpenAI API Docs
- OpenAI API changelog · OpenAI API Docs
- Model selection · OpenAI API Docs
- Using GPT-6 · OpenAI API Docs
- Codex models · OpenAI · ChatGPT Learn
- Codex pricing and usage limits · OpenAI · ChatGPT Learn
- ChatGPT and Codex changelog · OpenAI · ChatGPT Learn
- Speed: Fast and Ultrafast modes · OpenAI · ChatGPT Learn
- Reasoning models: reasoning effort · OpenAI API Docs
- Codex rust-v0.161.0 · GitHub · openai/codex ·
- Day 7: GPT-6.1 Sol default context 500k in Codex and Work · X · @thsottiaux ·
- No long-context fees for GPT-6.1 Sol on the subscription · X · @thsottiaux ·
- Artificial Analysis: GPT-6.1 Sol model page · Artificial Analysis
- Artificial Analysis on Muse Wonder: Intelligence Index, cost per task, Coding Agents Index · X · @ArtificialAnlys ·
Last checked . Facts change; the sources above are authoritative. See how we verify or report a correction.
