GPT-6 Astra: What’s Actually New in OpenAI’s New Frontier Model

OpenAI has released GPT-6 Astra, its newest frontier model, less than a week after Anthropic’s Claude Fable 5.1. OpenAI calls Astra the world’s most intelligent and aligned model yet. What actually makes Astra different? The simplest way to put it is this: Astra is built to do more, not just answer more. It can use a computer to complete tasks instead of telling you how to do them. It can create finished documents instead of giving you a rough first draft. During long coding sessions, Astra can also remember what happened earlier instead of starting from scratch. And it’s better at knowing when to take action and when to stop and ask. The bigger shift isn’t just smarter answers. AI is getting closer to actually getting the work done. Some important caveats remain, though. Astra doesn’t beat every competing model, and some of its biggest benchmark numbers come with asterisks. What Actually Changed in GPT-6 Astra Drive a Computer End-to-end Astra can fill out forms, update CRM records, run frontend QA checks on a website, and troubleshoot software by watching what happens on screen, without step-by-step hand-holding. OpenAI reports a score of 72.6% on OSWorld 2.0, narrowly ahead of Claude Opus 5’s 70.2% and well ahead of GPT-5.6 Sol’s 65.7%. Example: GPT-6 Astra can extract information from a W-2 and enter the relevant details into a Form 1040, helping automate routine tax-preparation work while leaving the final return for human review. Knows When to Ask and When to Guess Earlier models tended to either guess wrong on ambiguous instructions or interrupt with unnecessary questions. Astra is trained to fill in routine gaps on its own and pause only when the answer would meaningfully change the outcome. OpenAI’s own side-by-side demo shows the difference: GPT-5.6 Sol built a personal career website on its own in about 13 minutes. Astra paused after 20 seconds to ask what career the user was actually moving into. That’s a small moment, but it illustrates the difference between a model that acts and a model that uses judgment about when to act. Produces Finished Documents Astra is trained to match your existing templates, tone, and structure, and to pull in only the context that’s relevant rather than padding the output with everything it knows. Astra builds a slide deck from just a handful of template slides while keeping the tone and layout consistent throughout. Remembers Across Long Coding Sessions In Codex, Astra can now keep searchable notes across context windows instead of repeatedly compressing long debugging sessions into a single summary, preserving details that compaction tends to lose, like why an earlier fix failed. It’s opt-in for now through Codex’s config file, and OpenAI says it will become the default in the coming weeks. Cybersecurity Astra’s biggest jump isn’t a productivity feature at all. OpenAI reports Astra reaches the "Critical" threshold for cybersecurity under its own Preparedness Framework, its highest risk tier, meaning the model can independently identify and develop working exploits for previously unknown vulnerabilities. OpenAI reports a 100% score on ExploitBench and says Astra solved 88% of SRE-Bench reverse-engineering tasks on the first attempt. Because of that risk tier, OpenAI is gating the more dangerous parts of this capability at launch. Astra will help with defensive work like secure code review and patch validation, but it refuses to create proof-of-concept exploits until OpenAI expands access through its Daybreak program. GPT-6 Astra Benchmarks All figures are self-reported by OpenAI in its launch materials. Benchmark GPT-6 Astra GPT-5.6 Sol Claude Fable 5.1 Claude Opus 5 Gemini 3.8 Flash OSWorld 2.0 (computer use) 72.6% 65.7% – 70.2% – FrontierMath Tier 4 97.6% 83.0% 87.8% 73.2% – GPQA Diamond 96.0% 94.6% 93.7% 93.7% 95.3% Terminal-Bench 4.0 (coding) 57.7% 37.3% 55.8% 52.3% 19.1% ExploitBench 100.0% 78.5% – 70.0% – Humanity’s Last Exam (w/ tools) 57.2% – 65.0% 63.6% – What this actually tells us: Astra’s clearest strength is computer use. It leads Claude Opus 5 on OSWorld 2.0 and does it in meaningfully less time per task. Its coding scores have improved substantially over its own predecessor, though the lead over Claude Fable 5.1 on Terminal-Bench 4.0 is narrow rather than decisive. It is not universally ahead of Claude. On Humanity’s Last Exam with tools, Astra scores 57.2%, behind both Claude Fable 5.1 (65.0%) and Claude Opus 5 (63.6%). Some headline numbers rely on evaluation setups that don’t reflect normal usage. Astra’s marketed 99.9% on ARC-AGI-3 depends on an expensive, stateful evaluation harness. Independent testing by the ARC Prize Foundation found that a standard, stateless API call scores far lower, somewhere between 17% and 63% depending on the reasoning tier used. GPT-6 Astra vs Claude Fable 5.1 Comparison As neither model is publicly available yet, early access users on X have been sharing interesting examples. Building a 3D Villa from a Single Prompt Developer Karan Kendre posted a head-to-head Blender comparison, asking both models to build the same villa scene. Astra comes out ahead, with a more polished and realistic-looking result, particularly in interior details. Designing a Travel App from the Same Prompt Both models were prompted to design an app with the same goal. Fable went sky-themed, while Astra took an astro-inspired approach. Astra comes out ahead, with a more distinctive visual identity and stronger thematic consistency. Video Prompt Generation Both models generated a prompt for the same video concept, then created using Higgsfield AI. Astra comes out ahead, with a more coherent sequence and stronger visual storytelling. Cost of GPT Astra GPT-6 Astra is rolling out in stages: limited organizations first, then all ChatGPT Plus, Pro, Business, and Enterprise users over the following days. For developers, Astra is available as gpt-6-astra through the OpenAI API, Microsoft Azure, and Amazon Bedrock. API pricing: Input: $10 per million tokens Output: $50 per million tokens Fast mode: roughly 2.5x the speed at 2x the price That’s well above GPT-5.6 Terra’s $2/$12 rates and Claude Opus 5’s $5/$25 rates. The pricing is a signal: this is built for high-value autonomous work, not routine bulk-text generation. The model also supports zero data retention for eligible API customers. Conclusion The real test for Astra won’t be whether it can top another benchmark. It will be whether companies can hand it a messy, multi-step task and trust it to get the job done with minimal supervision. On paper, Astra looks like a significant step forward, particularly in computer use, coding, and agentic workflows. But it doesn’t lead everywhere. Claude’s models still have an edge on broader reasoning benchmarks. The bigger question is what happens outside the leaderboard. Once Astra is widely available, real-world testing will show whether its strengths translate into reliable, everyday workflows. That’s the test that matters most.