OpenAI launches GPT-6 Astra: 99.9% on ARC-AGI-3, from $10/$50 per 1M tokens
reach_vb · x · 2026-09-04
OpenAI introduces GPT-6 Astra, billed as its most intelligent and aligned model yet:
Capabilities and scores
- State-of-the-art computer use and software engineering, new breakthroughs in math and science, polished documents/spreadsheets/presentations that follow your templates
- 99.9% on ARC-AGI-3, 98% on FrontierMath Tier 4, and has already helped solve long-standing open math problems
- With an improved Codex harness, it completes tasks 1.9x faster than GPT-5.6 Sol on Mind2Web
Agents and safety
- In Codex, Astra can ask questions while continuing independent work; an experimental context feature lets it keep notes and search earlier context windows across long tasks
- Stronger cybersecurity capabilities backed by additional safeguards and monitoring against unauthorized actions
Availability and pricing
- Rolling out first to a limited set of organizations (including the Daybreak Access program); broader availability across ChatGPT Plus/Pro/Business/Enterprise, the OpenAI API and AWS in the coming days
- API pricing per 1M tokens: Standard $10 input / $50 output; Fast $20 input / $100 output for up to 2.5x the speed
Related event: OpenAI Launches GPT-6 Astra, Declaring 'Welcome to the AGI Era'(89 posts)→
More from coding & agent
- The fastest way to shortlist AI agent use cases: target the gaps between systems — kashifmanzoor · 2026-09-04
- Prime Intellect's RL stack adopts NIXL, cutting 800B-model weight transfer 9x from 86s to under 4s — xeophon · 2026-09-04
- GPT-6 Astra: the real leap is letting AI operate your computer — talkaboutdesign · 2026-09-04
- Alex Zhang defines the agent harness: code around an LLM, an agent once placed in an environment — CShorten30 · 2026-09-04
- OpenAI's Heidel: compaction in Responses API lets GPT-6 Astra max out ARC-AGI-3 — stevenheidel · 2026-09-04
- Non-engineer at Earendil shares how he customized the Pi coding agent for daily work — solyarisoftware · 2026-09-04