AI Daily · Sep 30, 2026: OpenAI launches GPT-6.1 Sol, dots agents, and Pro 500 plan in one day
OpenAI released GPT-6.1 Sol at one-fifth the price of Astra with a 95% cache read discount, alongside always-on dots agents, a Pro 500 plan with 25x Plus limits, and an Ultrafast 8x speed tier, available across ChatGPT Work and Codex.
Pick a date
-
OpenAI releases GPT-6.1 to paid plans and API
OpenAI announced GPT-6.1, describing it as highly efficient, now included in all paid ChatGPT plans and available on the API.
-
OpenAI launches Pro 500 plan with 25x Plus limits and Ultrafast tier
OpenAI announced Ultrafast, a premium speed tier delivering up to 8x faster token generation (300 tokens per second) in Codex and up to 6x in the API, available today for GPT-6 Astra with GPT-6.1 Sol coming soon; a new Pro 500 plan with 25x Plus usage limits and Ultrafast access is being introduced, and Pro 200 subscriptions are reopening with access to Astra and the new GPT-6.1 Sol model.
-
OpenAI launches dots agents and ChatGPT Space collaboration tools
OpenAI announced dots, always-on agents powered by GPT-6 Astra that can be configured with boundaries and run on their own cloud computer, plus ChatGPT Space, a collaborative workspace with interactive pages. Dots will be available in ChatGPT on web, mobile, and desktop for Pro, Business Premium, and Enterprise users in eligible markets.
-
ChatGPT subscription usage now works across 60 plus partner products
OpenAI's Tibo announced that ChatGPT subscribers can now use their included usage directly within over 60 partner products, including Devin, OpenCode, and Lovable, by signing in with ChatGPT.
-
Codex Security Cloud adds default cyber model access and repo scanning
OpenAI announced Codex Security Cloud is receiving a major upgrade, including default access to cyber-capable models via Daybreak Blue, continuous scanning of full GitHub repositories and new commits, automatic deduplication and fix preparation, available as a plugin in Codex desktop and web.
-
Requesty gateway adds new OpenAI model to its 600-plus model lineup
Requesty announced that a new OpenAI model is now live on its AI gateway, which offers unified API access to over 600 models.
-
Open models GLM-5.3 Flash and Kimi K3 available in Codex
Enterprise teams can now use open models like GLM-5.3 Flash and Kimi K3 natively within Codex, with usage counted against their OpenAI commit.
-
OpenAI introduces Decisions API for sub-300ms constrained decisions
OpenAI announced a Decisions API powered by Luna for fast constrained decision-making, supporting visual inputs and tuned to complete decisions end-to-end in under a few hundred milliseconds.
-
Anthropic launches Interviewer study following 81,000-person survey
Anthropic launched a new study using Anthropic Interviewer to gather user experiences and expectations about AI, building on last December's qualitative survey of 81,000 people; publishing interview results is optional.
-
GLM-5.3 uncovers thousands of vulnerabilities in open-source projects
GLM-5.3 has helped defend 389 open-source projects by finding 4,249 potential vulnerabilities so far through the still-running free OpenVuln service, with results shared privately with maintainers.