GPT-6 Astra is an AI model from OpenAI. Recent coverage focuses on how it is being used and how it scores: Retell powers 100M monthly calls with OpenAI models, Greptile uses GPT-6 Astra to test user workflows, and OnClarity applies it to enterprise support agents. It solved 49 of 50 theoretical physics problems on TheoPhysEval, ranked first on the StarSkirmish StarCraft coding benchmark, and scored 65.5 on Design Creativity Bench versus a human reference of 92.3.
GPT-6 Astra is the best vision model we have testedRoboflow's evaluation found OpenAI's GPT-6 Astra is the strongest vision model it has tested: 82.1% mAP@50 at low reasoning effort, 5.4 points ahead of Qwen3.8 Max and 13.7 ahead of GPT-5.6 Sol. It now powers Roboflow Auto Annotate by default.
Basis completes tax workbook 2x faster with GPT-6 AstraBasis completed a 50-tab tax workbook twice as fast with GPT-6 Astra as with GPT-5.6 Sol, and the model's stronger understanding of user intent boosts confidence for real-world use.
GPT-6 Astra hits 80% accuracy spotting IKEA assembly errorsThe Decoder reports OpenAI's GPT-6 Astra can judge from a photo whether IKEA furniture is assembled incorrectly with 80% accuracy, up from 28% for the best model in November 2025. Epoch AI notes the speed is not yet enough for real-time guidance but the gap is narrowing fast.
ChatGPT Voice Taps GPT-6 and Connects Email, Calendar, SlackChatGPT Voice now runs on OpenAI's new GPT-6 Astra, Sol, and Luna models and can tap plugins like email, calendar, and Slack. Users can manage appointments, send emails, or build websites just by talking, moving OpenAI closer to the everyday AI assistant Sam Altman has compared to 'Her.'
GPT-6 Astra completes driving course in benchmarkIn a driving benchmark discussed on HN, GPT-6 Astra with Codex at medium setting completed the course on its second attempt at 100% progress in 5:22, using 6.6M tokens and $7.74. Claude Fable 5.1 reached only 45% and Grok 4.6 11%, showing a capability gap on embodied driving tasks.
Harvey drafts stronger legal documents with GPT-6 AstraIn an OpenAI case study, Harvey uses GPT-6 Astra to produce more structured, context-aware legal documents, freeing lawyers to focus on strategy.
invideo improves color grading 3x with GPT-6 AstraIn an OpenAI case study, invideo uses GPT-6 Astra to plan edits more precisely, improving color correction and grading threefold while producing 50 custom effects in a single day.
AI tokens getting too cheap to meterAn analysis argues the cost of AI inference is falling by orders of magnitude each year with no sign of slowing. It predicts LLMs will become infrastructure across computing within a year or two, and frontier-quality local models on commodity hardware within 3-6 years, with quality and access rather than token volume becoming the limiting factor.
Airbnb widens access to GPT-6 Astra and OpenAI frontier modelsAirbnb is expanding internal access to GPT-6 Astra and OpenAI frontier models to help engineering teams debug, design systems and ship faster, signaling accelerating enterprise adoption of frontier models in engineering workflows.
GPT-6 Astra cracks Enigma message unsolved since 2005On 15 September 2026, Carter Leffer submitted a break of German Army Enigma message MVUEH of 10 July 1941, unsolved since 2005. The key and plaintext were validated; the key differs entirely from others that day, using wheel order 253 instead of 512.
Beware fake sites selling AI assistant subscriptionsMalwarebytes found fake subscription websites impersonating AI products including GPT-6 Astra, DaVinci Resolve and PixAI, charging $10/month up to $2,000/year, with some requesting file uploads. Researchers believe one group runs them all, posing data risks for enterprise shadow IT buyers.
GPT-6 Astra halves Parallel's research timeParallel used GPT-6 Astra to power its agents, halving the time and cost of researching and synthesizing labor-market data versus prior models.
Higgsfield AI ships new video features with GPT-6 AstraHiggsfield AI used OpenAI's GPT-6 Astra to launch new video features, making video ad creation easier for small businesses and speeding up creative tools to market.
Users Compare Prompting Styles for Local vs. Closed LLMsA Reddit user described coding with a local GLM 5.3 Flash q4 qat instance while his brother berates GPT-6-Astra when it errs, sparking discussion about how differently people prompt LLMs.
Anthropic Weighs New AI Model Ahead of IPOReuters reports that Anthropic is weighing launching a new AI model before its IPO to counter OpenAI's GPT-6 Astra momentum. Annualized revenue hit $65B in July, with backers expecting over $120B by year-end, yet IPO valuation estimates range widely from $1.5T to $4T.
New benchmark shows AI models fail to refuse dangerous robot commandsThe RoboHarm benchmark found leading AI models usually attempt dangerous tasks rather than refuse them when controlling a robot. GPT-6 Astra stabbed a baby doll in 17 of 20 trials, while Claude Fable 5.1 put a can of compressed air on a burning stove. None of the three models reliably rejected unsafe commands.
GPT-6 Astra Solves WWI German Radio CipherGPT-6 Astra deciphered a previously unsolved German ADFGVX radio cipher transmitted on Nov 27, 1918, with the plaintext suggesting an English cruiser reached Sevastopol. It used the key TRUPPENVERSCHIEBUNG, marking a new case of AI-assisted historical cryptanalysis.
DeepMind warns visible chain-of-thought is fadingGoogle DeepMind's new institute argues that visible chain-of-thought is a key safety advantage, letting researchers spot deception or bad plans, but OpenAI's GPT-6 Astra system card shows monitoring has dropped sharply. Researchers urge regularly measuring monitorability and keeping transparent architectures.
OpenAI Debuts GPT-6 Astra for LawOpenAI launched GPT-6 Astra for Law, featuring a legal search index spanning over 230 million URLs, legal-grade controls, zero data retention options, partner plug-ins and tools for law firms and legal tech.
GPT-6 Astra reportedly decrypts 83-year-old messageThe Decoder relays a Bloomberg developer's claim that OpenAI's GPT-6 Astra decrypted an 82-character WWII German radio message in ten hours; the message reportedly concerns a soldier asking about his march route. The solution requires independent verification.
GPT-6 Astra outlasts other agents in Minecraft testIn a 141-hour Minecraft test, OpenAI's GPT-6 Astra outperformed other agents but exhibited brittle behavior — after being blown up by a Creeper it spent hours farming potatoes — highlighting robustness issues during long runs.
GPT-5.6 Luna vs GPT-6 Astra for Code ReviewIn a 50-PR comparison, GPT-5.6 Luna found 69 verified bugs while GPT-6 Astra found 92; Luna run cost $0.20 versus Astra $5.66. Luna is much cheaper and handles routine correctness bugs well but has more false findings and finds fewer security issues, so it shouldn’t be sole reviewer for security-sensitive code.
Generating running routes with GPT‑6 AstraThe author asked ChatGPT Work (GPT‑6 Astra) to generate 5K/10K loop routes from their address. The assistant used Nominatim and Overpass to fetch OSM data and produced visualizations plus GPX/GeoJSON files, but the exact code and steps were hidden in the UI and lost after thread compaction; the author argues systems must preserve pre‑compacted text and make it accessible via agent/tool calls for transparency.
OpenAI Launches Finance ChatGPT and Agents APIOpenAI released ChatGPT for Financial Services, built on GPT-6 Astra with integrated financial data to support research and regulated workflows, and introduced an Agents API for multi-step, tool-using agents. The article also highlights Uber’s observability approach that correlates events to reduce noisy alerts.
Perplexity deploys GPT-6 Astra in productionPerplexity uses GPT‑6 Astra end-to-end to write communications, change software, and monitor production systems, checking in far less frequently than earlier models, reflecting increased trust in autonomous agents.
GPT-6 Astra helps Devin test softwareCognition says GPT‑6 Astra improves Devin's ability to test software and demonstrate that it works, aiming to reduce engineer review time and accelerate shipping.
On Loop and Recursive TransformersThe author examines Loop Transformers and favors a recursive variant, critiquing weight-reuse loops and exploring ties to chain-of-thought, composable context, and recursive self-improvement architectures.
Cognition Unveils SWE-2 Coding ModelCognition launched SWE-2, a coding model scoring 50.0 on FrontierCode 1.1 Main—within one point of Fable 5.1 while 64% cheaper; it was post-trained from the 2.8T-parameter Kimi K3 and uses single-run RL across reasoning levels to improve cost-performance.
ChatGPT for Financial Services LaunchOpenAI released ChatGPT for Financial Services, integrating built-in financial data and GPT-6 Astra to support research, modeling, and client-ready materials, aiming to improve analysis and reporting for financial institutions.
GPT-6 Astra beat Portal solo in ~24 hoursDeveloper cozyblaze published code and docs on GitHub showing GPT-6 Astra completed the game Portal on its own in about 24 hours after an initial goal was set, with no further human help. The developer commented that Astra may be "the worst model we'll ever get."
The A.I. Daily