OpenAI DevDay drops 20 launches: always-on agent dots, GPT-6.1Sol, cloud Codex, Pro500 tier
新智元 · wechat · 2026-09-30
Altman unveiled 20+ launches at OpenAI DevDay, recasting ChatGPT from Q&A tool to do-the-work platform.
dots: OpenAI's first 7x24 always-on agent with its own cloud computer and browser, 4000+ tool integrations, lives in Slack/Teams; keeps running tasks with the screen off. Demos included autonomous site building/launch, updating a board deck from Teams feedback (revenue +51% YoY), drafting migration architecture from meeting notes, and auto-preparing a forgotten invoice. Passwords aren't exposed to the model; idle dots only do read-only background research. Available to Pro/Business Premium/Enterprise.
ChatGPTSpace: an AI-native collaboration space where humans and dots share one knowledge base and project context; @ a dot to generate charts or run analysis.
GPT-6.1Sol: matches Astra on coding, beats Opus 5.5 on long-horizon automation and science reasoning, at 1/5 Astra's price with $0.10/M cache reads. Ultrafast mode: 300 tokens/s, 8x speed. Pro $200 returns, plus new Pro500 tier with 25x Plus usage and Ultrafast access.
Codex goes cloud: preconfigured reusable dev environments, phone remote control, voice-prompted CLI, /agents multi-task view, desktop code review pushed to GitHub/GitLab, and Codex Security Cloud for repo-wide vulnerability scanning.
APIs: the tech behind dots ships as Agents API (computer use, multi-agent orchestration, tool search, context compression); Decisions API handles fixed-option decisions (76/78 correct, 230ms latency). Responses API traffic up 100x YoY at 99.9% uptime.
Context: Meta Muse hit 3M downloads in three weeks, Google Spark already works around the clock — personal agents are the next battleground.
Related event: OpenAI Launches Dots, Always-On Agents Powered by GPT-6 Astra(112 posts)→
More from Models
- You don't need frontier pricing: DeepSeek and GLM flash models can do 90% of your work locally — PMinervini · 2026-09-30
- User claims GPT-6.1 Sol ULTRA ran 25 minutes on just 1% of weekly quota (unverified) — steipete · 2026-09-30
- GPT 6.1 Sol launches with 50% cheaper caching; Luna can also make motion videos — oran_ge · 2026-09-30
- Users report Grok Bot is now nearly as fast as Muse — yunta_tsai · 2026-09-30
- banteg: model recovers 88 decompiled functions per hour that recompile to exact original machine code — banteg · 2026-09-30
- Will AI subscriptions hit $1,000/month next year? Priced against a $200k salary, some say it's cheap — RachelVT42 · 2026-09-30