GPT-6 Sol vs GPT-6.1 Sol: differences and Codex limits

Quick answer: GPT-6.1 Sol is the newer Sol model and the one OpenAI recommends for complex coding in Codex. Both cost $2 per million input tokens and $10 per million output tokens, with the same 1,050,000-token context and 128K max output. GPT-6.1 Sol halves the cached-input price to $0.10, has a later knowledge cutoff (30 April against 20 April 2026), drops the none reasoning setting and is the only one of the two with Ultrafast mode. In Codex on Plus and Business, OpenAI estimates 15 to 160 local messages per five hours with GPT-6.1 Sol and 15 to 150 with GPT-6 Sol.

Key facts

• Released in the API: GPT-6 Sol on 22 September 2026, together with GPT-6 Luna; GPT-6.1 Sol on 29 September.
• Same price: $2 input and $10 output per million tokens, $2.50 cache writes. Cached input: $0.20 on GPT-6 Sol, $0.10 on GPT-6.1 Sol.
• Same size limits: 1,050,000-token context, up to 922,000 input tokens and 128,000 output tokens.
• Knowledge cutoff: 20 April 2026 for GPT-6 Sol, 30 April 2026 for GPT-6.1 Sol.
• Reasoning effort: GPT-6 Sol supports none, low, medium (the default), high, xhigh and max. GPT-6.1 Sol has the same list without none.
• Ultrafast, at 6x the Standard price: listed for GPT-6.1 Sol, not for GPT-6 Sol.

Side by side in the API

OpenAI's model page describes GPT-6 Sol as built for complex coding and agentic workflows and points to GPT-6.1 Sol as the newer Sol model. OpenAI pitches GPT-6.1 Sol as near-Astra performance for complex work at a lower cost. Both bill prompts over 272K input tokens at 2x the input and cache rates and 1.5x output for the whole request. On both, Batch and Flex cost half and Fast mode costs double. Ultrafast, a faster service tier that OpenAI added to the Responses API for GPT-6.1 Sol on 8 October, costs 6x Standard. GPT-6 Sol's page lists no Ultrafast tier.

Codex limits: Sol, 6.1 Sol, Astra and Luna

OpenAI's Codex pricing page estimates local messages per five-hour window. The ranges are the same on Plus and Standard Business:

• GPT-6.1 Sol: 15 to 160.
• GPT-6 Sol: 15 to 150.
• GPT-6 Astra: 5 to 45.
• GPT-6 Luna: 350 to 3,000.

Pro plans currently have no five-hour limit, and weekly limits may also apply. OpenAI says these are estimates, not fixed message limits, and that local messages and cloud chats share the same allowance.

Which one Codex uses

• Codex CLI 0.161.0 (7 October) made GPT-6.1 Sol the default model in the bundled and Amazon Bedrock catalogs.
• GPT-6.1 Sol is on Plus, Pro, Business, Enterprise and Edu in the desktop app and CLI. On Enterprise and Edu it stays off until an admin enables it. Free and Go weren't included at launch.
• Since GPT-5.5 left Codex for ChatGPT sign-ins on 14 October, OpenAI names GPT-6.1 Sol as its replacement on paid plans. See our post on the "gpt-5.5 is not supported" error.
• Ultrafast with GPT-6.1 Sol in Codex and ChatGPT Work is on Pro $500 and eligible Enterprise and Edu plans.
• To pick one: /model in the CLI, codex --model gpt-6.1-sol (or gpt-6-sol), or model = "gpt-6.1-sol" in ~/.codex/config.toml.
• OpenAI's Codex docs list both Sol models and Luna as Work and Codex models, not Chat models. The Chat tab's GPT-6 is covered in our GPT-6 in ChatGPT post.

Where GPT-6 Luna fits

GPT-6 Luna is the low-cost model in the same family: $0.10 input, $0.01 cached input and $0.50 output per million tokens, a 1,050,000-token context, a knowledge cutoff of 18 May 2026 and reasoning from none to max. In Codex it has by far the most room, 350 to 3,000 local messages per five hours on Plus and Business, and it's the model Free and Go get in the desktop app. Codex lets Luna reason up to Max, but not Ultra.