OpenAI’s September 29, 2026 post introduces GPT-6.1 Sol as an upgrade to GPT-6 Sol, not a new family. It “nearly matches” GPT-6 Astra on agentic coding, computer use, and professional work at one-fifth of Astra’s standard input and output token prices. For Alberta teams already running agents on the earlier Sol tier, the change is a price and capability step, not a reason to throw out the workflow. Our earlier briefing is GPT-6 Sol pricing for Canada.
What OpenAI says the upgrade is
OpenAI’s product post is the primary. It describes GPT-6.1 Sol as an upgrade that nearly matches Astra on three kinds of work, at one-fifth of Astra’s standard input and output prices. Cached input is $0.10 per million tokens, which the post says is 95% less than standard input pricing and 50% less than GPT-6 Sol’s cached input pricing. The post does not say the model is generally available in ChatGPT. It says it is “available starting today” to Plus, Pro, Business, Enterprise, and Edu users in ChatGPT Work and Codex, and that it is “not yet available in Chat.”
“We’re introducing GPT‑6.1 Sol, an upgrade to GPT‑6 Sol that nearly matches GPT‑6 Astra’s intelligence on agentic coding, computer use, and professional work at one-fifth of Astra’s standard input and output token prices.”OpenAI — Introducing GPT-6.1 Sol, September 29, 2026
The DevDay 2026 recap, also dated September 29, 2026, shortens availability to “all API, Plus, Pro, Business, Enterprise, and Edu users.” It does not repeat the product page’s line that the model is not yet in Chat. This briefing follows the product page on that point. Neither page uses the words “general availability” for ChatGPT or the API.
API price, context, and the long-prompt multiplier
Developers call it as gpt-6.1-sol. The product page states standard API prices of $2 per million input tokens, $0.10 per million cached input tokens, and $10 per million output tokens. Re-checking the API model card on October 2, 2026, the card prints a 1,050,000 context window and 128,000 max output tokens, with text token prices of $2.00 input, $0.10 cached input, and $10.00 output per 1M tokens. The same card prints cache writes at $2.50 per 1M, and it says prompts with more than 272K input tokens are priced at 2x input and cache rates and 1.5x output for the full request. Fast mode is 2x standard. Batch and Flex are 50% lower than standard. Regional processing adds a 10% premium where available. Those multipliers are on the card. They are not in the short product-post price sentence.
| Token | Standard | Note on the card |
|---|---|---|
| Input | $2.00 | 2x above 272K input tokens, full request |
| Cached input | $0.10 | 5% of the uncached input rate |
| Cache writes | $2.50 | 1.25x uncached input |
| Output | $10.00 | 1.5x above 272K input tokens, full request |
OpenAI says GPT-6.1 Sol Ultrafast, with up to 8x faster token generation compared with its standard speed in Codex, comes “in the coming days.” The product page does not say Ultrafast is on today, and it does not give an API speed multiple for this model’s Ultrafast tier.
GitHub Copilot is a separate, narrower rollout
GitHub’s changelog, dated September 29, 2026, says GPT-6.1 Sol “is now generally available and rolling out in GitHub Copilot” for agentic coding and terminal workflows. That “generally available” sentence is GitHub’s, about Copilot, not a claim on OpenAI’s product page. GitHub lists Copilot Pro+, Max, Business, and Enterprise. It does not say every Copilot plan. The model is “billed at provider list pricing under usage-based billing.” Business and Enterprise administrators manage access in the model policy. Rollout is gradual. If the picker does not show it, GitHub says to check back.
What Alberta operators should budget now (our advice)
What follows is our advice, not OpenAI’s or GitHub’s. If an automation already runs on GPT-6 Sol, re-price one real trace on gpt-6.1-sol before you switch the default. Cached input at $0.10 per million is the number that matters for agents that resend the same context. A prompt over 272K input tokens is not the $2 rate. Read the card before you forecast a month of agent traffic.
Do not tell staff the model is in ordinary Chat. OpenAI says Work and Codex for the paid plans named above, and not Chat yet. In Copilot, confirm the seat is Pro+, Max, Business, or Enterprise, and that a Business or Enterprise admin has not disabled the model. For a contained pilot, AI consulting Alberta work should start from a workflow you already run, with a spend cap, not from a benchmark slide. An AI opportunity scan is a better first step than a blanket model swap.