OpenAI has launched GPT-6.1 and Sol with Codex and ChatGPT Work, claiming that their performance is close to that of Astra and the cost is lower.
Ai Focus
OpenAI announced during the Developer Day event that they have launched the GPT-6.1 Sol model on Codex and ChatGPT Work. According to the official statement, this model performs close to Astra when handling complex tasks, but at a lower cost than Astra. It is suitable for repetitive, long-running tasks in code, applications, and documentation. Based on information disclosed by @LuminaBench, the pricing is $2 per 1 million tokens inputs, $0.1 for cache inputs, and $10 for outputs.
On September 30th, at the OpenAI Developer Day event held today, OpenAI launched the GPT-6.1 Sol model in Codex and ChatGPT Work. Officials stated: "The GPT-6.1 Sol perform close to Astra when handling complex tasks, but at a lower cost than Astra."
The official introduction of IT's translation service is as follows:
GPT-6.1 Sol When handling complex tasks, its performance is close to that of Astra, but the cost is lower than that of Astra. It is suitable for repetitive, long-running tasks in code, applications, and documentation. Availability depends on the user's package, client settings, and workspace configurations.
In terms of pricing, according to the information revealed by @LuminaBench:
- The input fee for every 1 million tokens is 2 US dollars (Note from IT: According to the current exchange rate, this is approximately 13.4 yuan RMB);
- The caching input fee for every 1 million tokens is 0.1 US dollars (which is approximately 0.67 Chinese yuan at the current exchange rate);
- The output fee for every 1 million tokens is 10 US dollars (which is approximately 67.2 yuan according to the current exchange rate).
Follow CoinMeta official accounts to stay updated
Views 18
CoinMeta reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”，and we will handle it promptly。
No comments yet. Be the first!
Trump predicts that the war with Iran will end soon, while talks in Hormuz are still ongoingU.S. President Donald Trump stated that he expects the war with Iran to end "soon," following communications between U.S. officials and mediators pushing for a resumption of the agreement regarding the Strait of Hormuz. Meanwhile, both sides are still negotiating over the revised plan, but a ceasefire has not yet been announced; Trump also denied having proposed to relax sanctions or unfreeze funds in exchange for Iranian nuclear concessions.
crypto.news
·2026-09-30 03:22:3810 
The Ohio State University reaches a partnership with Google to accelerate scientific research and the construction of AI literacyThe Ohio State University and Google announce a new strategic partnership to establish a AI space within the university's innovation park, aiming to promote scientific research discoveries, expand the use of technology on campus, support the development of AI skills, and provide support for students from enrollment to graduation and beyond. The collaboration includes introducing Google DeepMind collaboration, providing access to Gemini Enterprise and Gemini Notebook throughout the university through Google for Education, as well as launching a student research ambassador program.
PR Newswire
·2026-09-30 03:22:359 
GTT Launches AI Native Network Defense PlatformGTT Communications announces the launch of GTT Defense Halo, a native network defense platform designed to proactively identify and fix issues before attackers discover vulnerabilities, as well as to real-time assess whether network anomalies pose a real threat. The platform is built on the AI infrastructure of GTT and deploys dedicated instances for each customer.
The Block
·2026-09-30 03:14:179 
OpenAI launches GPT-6 Astra Ultrafast service, with a maximum of 300 tokens per secondOpenAI launched a new service level, Ultrafast, at the 2026 Developer Day event. According to the official statement, GPT-6 Astra Ultrafast can achieve up to 300 tokens per second in Codex, with the generation speed increasing by up to 8 times in Codex and the call time increasing by up to 6 times; this service is now available for API. Pro 500 subscribers and enterprise users can experience it in ChatGPT Work and Codex.
The Block
·2026-09-30 03:14:1610 
Correction | MultiplierMultiplier releases a correction stating that it has achieved $100 million in annual recurring revenue ( ARR ), with an annual growth rate of 80%, and a core business net revenue retention rate of 115%. The company also stated that its North American business has become the main growth engine, with 60% of current revenue coming from North America, and the annual cross-border payroll processing volume exceeding $1.81 billion.
PR Newswire
·2026-09-30 03:02:4011 View More
