ModelsUpdated 2 min read5 sources
OpenAI releases GPT-6.2 Astra and Sol; Astra’s token prices are 20% below GPT-6 Astra’s
Sol keeps GPT-6.1 Sol’s prices; choosing a two-hour or eight-hour background window in the Responses API cuts token rates by 30% to 40%.
At a glance
- Date
- Topic
- Models
- Reading time
- 2 min
- Sources
- 5
This entry is compiled automatically with AI and is not read by a person before it is published. Its sources are below. If you see an error, write to us.

OpenAI announced GPT-6.2 Astra and GPT-6.2 Sol; the API changelog lists both under 17 November. According to the announcement they became available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex; on Enterprise and Edu a workspace admin has to enable them. Developers can also use gpt-6.2-astra and gpt-6.2-sol in the API.
- Astra, per million tokens: $8 input, $0.40 cached input, $40 output.
- Sol, per million tokens: $2 input, $0.10 cached input, $10 output.
OpenAI says Astra’s input and output token prices are 20% lower than GPT-6 Astra’s and its cached-input price is 60% lower, and that Sol keeps GPT-6.1 Sol’s prices. These are Standard API rates; according to OpenAI’s model pages, for requests with more than 272K input tokens, input and cache rates double and output costs 1.5 times the short-context rate for the full request.
The changelog says the Responses API public beta now offers a choice of background completion windows, two hours or eight hours. Compared with Standard rates, the two-hour option costs 30% less in tokens and the eight-hour option 40% less for eligible background requests. The option is listed for the two new models and for gpt-6.1-terra and gpt-6-luna.
The same changelog flags two behaviour changes in GPT-6.2: reasoning is handled differently when the start of a conversation is edited, and the top-level instructions field is treated differently. Tool calling needs the Responses API.
OpenAI’s announcement page adds that the models launch with tighter safeguards for higher-risk cybersecurity help; these lower the risk of misuse but also get in the way of some legitimate vulnerability research and security testing. Prices are in US dollars; no Turkish lira price is given.
What it means for you
The studio’s noteFor batch jobs where nobody waits for the result, such as overnight tagging or summarising, the background window looks like a way to lower the token bill. Before switching models, test how your code handles the instructions field and reasoning.