context is everything

Today, Trata is excited to share the revised Hedge-Bench: our benchmark tying open-ending reasoning to prediction capabilities. We constructed 92 real-world environments derived from hedge fund analysts’ reasoning traces. Agents reason from frozen information snapshots, and their forecasts are evaluated against future real-world outcomes. No model scores above 50% (GPT-6 Astra being the top performer at 44.2%). For frontier, state-of-the-art models, these results highlight a persistent and wide gap between producing plausible-sounding reasoning and making accurate, economically valuable forecasts. In a semi-verifiable domain like finance, where outcomes are delayed and reasoning is difficult to verify, intermediate expert reasoning traces are crucial. We post-trained @Meta's Muse-Glimmer-30B on 165 environments with per-turn credit assignment using an LLM-as-a-Judge. We observe evidence of generalization: held-out forecasting accuracy at high effort improved by 10.06 percentage points, while out-of-distribution evaluation scores improved by 3.94 points on @OpenAI's GDPval, 2.29 on @Surge’s GDP.pdf, 7.07 on @Harvey’s LAB, and 4.96 on @Samaya_AI’s FrontierFinance under their respective metrics.

@OpenAI DevDay Berlin was awesome. thank you everyone for coming! & special thanks to the person that gave me KitKat Club tickets as a present, it’s not just at DevDay that Berlin loves harnesses 🦮

Starting today, eligible ChatGPT Pro and Business subscribers can claim up to $500 per month in API credits to build apps and agents.

DevDay Exchange Berlin was a vibe! We had a great set of devs come out and spend the day in great conversations And closed with a punchy AMA with @romainhuet @whoiskatrin and @paw_lean ❤️ for @reach_vb, @kagigz, @dkundel

i'm afk one day and then we disprove ETH??? gonna close Codex again and see what happens to P vs NP

Berlin yesterday. Paris tomorrow. London next week 🇩🇪🇫🇷🇬🇧 Had a lot of fun demoing with our Applied AI Engineering crew and getting into what people are building. We’ll be speaking and demoing in Paris Oct 28 and London Nov 3. Come find us if you’re joining! @aloglute @wulfie_bain_ @HennersBro98

GPT-6.1 Terra is now available in Strawberry 1.0. It feels amazing. Fast. Smart. Cheap. Try it now.

@OpenAI DevDay Berlin: come find me lurking and speaking

We’ve been trying out OpenAI’s Decisions API at Lovable and really like what we’re seeing.

Excited to kick-off our DevDay Exchange in Berlin & Germany is cooking with Codex!! 🇩🇪 - weekly active users are up 20× since January - nearly 70% use the desktop app - 1 in 5 desktop users use skills for docs, slides and spreadsheets and Germany is now top 5 globally for active API developers. kudos to all of our community members and ambassadors for all their support and love. come find me after my talk and let’s hang!! 🚀

GPT-6.1 Terra: delivering fast and capable computer use at a lower price. GPT-6.1 Terra understands what’s on screen, takes action across apps, writes naturally, and is a more capable model for your whole workday with the efficiency to take on more of it.

AHHHH YES I LOVE YOU OPENAI QUICK @emilahlback run the benchmarks on real enterprise tasks!!

Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications. On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments. While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software. After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely. openai.com/index/securing-ou…

Hiring in Singapore 🇸🇬 for our Applied AI for Startups team at @OpenAI Come join a team of ex-founder/CTOs, MLEs, Research Eng's, DSs. We work with frontier startups, and closely with Product & Research. Obsessed with startups, high agency, & technical? Apply or reach out

Our latest model advances bring stronger capabilities for building and migrating software in low-resource programming languages such as COBOL, Fortran, and Visual Basic. These languages power software that banks, governments, and businesses depend on. Progress here means understanding decades of business logic and helping developers modernize those systems while preserving what works. We’re building these capabilities around the challenges developers and enterprise customers bring us. If your team is working on these systems, tell us which languages and workflows we should tackle next. openai.com/index/low-resourc…

Already showing people what they can build with Codex? Applications to become a Codex Ambassador are open again. We support you in hosting meetups or sharing what you’ve learned: developers.openai.com/commun…