--- title: "GPT-6 Astra hits 99.9% on ARC-AGI-3 — on a harness it built itself" description: "OpenAI shipped GPT-6 Astra on 3 September 2026 with a near-perfect ARC-AGI 3 score. Measured on the benchmark's own standard harness, the same model drops to 62.7% — and costs more." canonical: https://inyourgeek.com/en/articles/2026-09-03-gpt-6-astra-99-9-sur-arc-agi-3 type: article author: "Sébastien Soulier" publisher: "InYourGeek" language: en datePublished: 2026-09-03 dateModified: 2026-09-04 section: "AI" tags: ["openai", "benchmark", "llm"] --- # GPT-6 Astra hits 99.9% on ARC-AGI-3 — on a harness it built itself > OpenAI shipped GPT-6 Astra on 3 September 2026 with a near-perfect ARC-AGI 3 score. Measured on the benchmark's own standard harness, the same model drops to 62.7% — and costs more. September 3, 2026 · AI · 3 min ## The headline number, and its asterisk OpenAI put GPT-6 Astra online on 3 September 2026, first for a limited set of organisations, with a rollout to ChatGPT Plus, Pro, Business and Enterprise accounts, plus the API and AWS, promised for the days that follow ([the announcement](https://openai.com/index/gpt-6-astra/)). API pricing: $10 per million input tokens, $50 per million output — exactly what Claude Fable 5 and 5.1 charge. The entire competitive pitch, then, fits in a single row of a pricing table. The model will answer to `gpt-6-astra` once the rollout completes. The number doing the rounds since is 99.9% on ARC-AGI 3, a benchmark released in March 2026. Except that, as [Simon Willison](https://simonwillison.net/2026/Sep/3/gpt6-astra/) notes, the ARC-AGI blog publishes the protocol next to the score: those 99.9% cost $19,000 and were run on OpenAI's in-house harness, the Provider Adapter, which holds an opaque reasoning state between requests and compacts long conversations so the model can reuse its earlier work. On ARC-AGI's default harness, the same model manages 62.7%, for $26,000. Worse and dearer: what changed between those two numbers is not the model, it's the plumbing wrapped around it. And Claude Fable 5 has no published result on this benchmark at all, which leaves the implied comparison with nothing to stand on. ## The scores that need no special arrangements On security, the reported results are clean, and they read best against GPT-5.6 Sol: 100% on ExploitBench versus 78.5%, 42.4% on ExploitGym versus 30.3%, and 99.2% across four attempts on the binary reverse-engineering half of SRE-Bench, versus 68.7%. A genuine leap — as long as you also notice that ExploitGym is a test the model still flunks close to six times out of ten. Long context moves too: on OpenAI's own eight-needle benchmark, Astra posts 100% between 256,000 and 512,000 tokens, and 96.3% between 512,000 and a million. That is the sort of figure that, if it holds up outside the lab, quietly retires a problem everyone has been papering over with RAG for two years. ## The leaderboard Astra doesn't win Artificial Analysis puts Astra at 61 on its Intelligence Index — level with GPT-5.6 Sol, five points behind Claude Fable 5.1 at maximum effort with fallback, and behind Meta's Muse Spark 1.3. A new generation that fails to beat the old one on the index is an odd thing to attach a fresh major version number to. The same shop's Coding Agent Index tells a friendlier story: at maximum effort Astra sits on the cost-efficiency frontier, costing roughly what Sol costs for two extra points, and under half Claude Fable 5's price per task at an equal score. That, not ARC-AGI, is where the sales argument actually lives. ## What the system card doesn't tell you yet The [system card](https://deploymentsafety.openai.com/gpt-6-astra) is out, but most of the numbers in circulation are OpenAI's own self-reported benchmarks, on which Astra beats Fable in the majority of cases. The post itself, titled "A new generation of intelligence," was hard to actually read on launch day: OpenAI's blog was throwing 500s, and readers fell back on a mirror. The Hacker News thread had no such trouble — 638 points and 369 comments within hours. Willison, who hasn't been given access to the model yet, declines to go further. Scoring 99.9% on equipment you brought yourself is still a score. What the next few days will settle is how much of it is left once somebody else is holding the stopwatch. ## Sources - [GPT-6 Astra](https://openai.com/index/gpt-6-astra/) — openai.com - [GPT‑6 Astra](https://simonwillison.net/2026/Sep/3/gpt6-astra/) — simonwillison.net _Article written with AI assistance from the cited sources, reviewed and approved by a human before publication._ --- Equivalent HTML page : https://inyourgeek.com/en/articles/2026-09-03-gpt-6-astra-99-9-sur-arc-agi-3 Published by InYourGeek — https://inyourgeek.com/en/a-propos Method and safeguards : https://inyourgeek.com/en/methode All articles : https://inyourgeek.com/en/archives