On 29 September OpenAI announced more than twenty things at DevDay in San Francisco: dots, always-on agents with a computer of their own; GPT-6.1 Sol, which OpenAI says comes close to Astra at a fifth of the price; ChatGPT Space and Pages for teams working alongside agents; Codex in the cloud, computer use in the Agents API, and a Pro 500 plan. We are collecting the launches in OpenAI's own words, what was shown on stage, what people tried and argued about in the first hours, and the night before, each in the poster's own words. Nobody in these bylines is on eChai, and every entry links back to where it was posted.
9 entries Clear
Cognition, the company behind Devin, putting 6.1 Sol in Devin within the hour, with its own benchmark and its own cost per task.
Cognition
1h
GPT-6.1 Sol is now available in Devin.
On FrontierCode 1.1, GPT-6.1 Sol scores 60.4%, in line with GPT-6 Sol (60.7%), but at a fraction of the cost. At medium reasoning effort, it costs $0.31 per task, 81% less than GPT-6 Sol at max effort.
Derya Unutmaz, an immunologist and professor, had 6.1 Sol early too, and filmed this from the keynote as it was announced.
Derya Unutmaz
1h
Next up is GPT-6.1 Sol, which I also had early access to for some time, and I am super excited about it! In my testing it is as good as GPT-6 Astra but faster and much cheaper. This will have a massive impact on building with Codex!
OpenRouter, which routes one API to four hundred models, listing 6.1 Sol on launch day.
OpenRouter
1h
GPT-6.1 Sol from @OpenAI is live on OpenRouter!
It nearly matches GPT-6 Astra on agentic coding, computer use, and professional work at one-fifth of Astra's price.
$2/M input and $10/M output, with cached input at $0.10/M, half of GPT-6 Sol's cached price.
Yuchen Jin, at Databricks and before that co-founder of Hyperbolic, on the arithmetic of paying for speed.
Yuchen Jin
1h
Fast: 2x faster, 2x the price.
Ultrafast: 8x faster, 6x the price.
Maybe we should roll out some Ultra-ultrafast OSS model endpoints?
GPT-6.1 Sol went live in the API a few minutes before the keynote began. OpenAI's launch page says it matches GPT-6 Astra on its real-codebase coding benchmark at about a fifth of the cost, at $2 per million input tokens, $0.10 cached and $10 output.
OpenAI
1h
GPT-6.1 Sol: near-Astra intelligence for a fifth of the price.
It’s the most cost-efficient model for its performance available today.
Tibo on the same launch, and the speed tier that came with it. OpenAI's recap puts Ultrafast at up to eight times the standard speed in Codex (300 tokens a second) and up to six times in the API, on Astra today and on 6.1 Sol soon.
Tibo
1h
Introducing 6.1 Sol, near Astra intelligence at one fifth of the price of Astra and 95% cache read discount. It is an absolute workhorse.
Combined with ultrafast for 8X speeded available today for Astra and coming soon for 6.1 Sol.
What Ultrafast costs on Astra in the API, posted with the price list the moment it appeared.
Leo
2h
GPT-6 Astra Ultrafast is an absolutely outrageous $60/$300 per Mtok btw
From the Hacker News thread on the GPT-6.1 Sol launch page. Anthropic released Claude Sonnet 5.5 the day before.
gradus_ad, in Hacker News
Ominous for the industry and investors that token price is becoming the main battleground. Could be Anthropic's rationale for IPOing this year.
Token Gremlin, reading the API's price list minutes after GPT-6.1 Sol appeared in it. The unchanged prices are the point OpenAI's own page makes: the same money buys a model it says is close to Astra.
Token Gremlin
2h
GPT-6.1 Sol pricing is showing up as:
$2/M input
$10/M output
$0.10/M cached input
$2.50/M cache writes
Looks basically unchanged from GPT-6 Sol pricing (unless I’m missing something).
Now I really want to see where the actual gains landed in the benchmarks.
