GPT-6.1 Sol
GPT-6.1 Sol is the middle tier of OpenAI's GPT-6 family, available via API and rolling out in ChatGPT Work and Codex. Recent coverage highlights its claimed 86.6% to 97.9% COBOL migration accuracy, strong political biases in election recommendations, and availability on Google Cloud. The model also appears in studies regarding human prioritization and medical reasoning benchmarks.
GPT-6.1 Sol is a coding model from OpenAI, first covered on 22 September 2026. It is available through an API. Its context window is 1,050,000 tokens.
Last updated
Quick facts
- Who makes GPT-6.1 Sol?
- GPT-6.1 Sol is made by OpenAI. OpenAI news →
- When was GPT-6.1 Sol released?
- The first report of GPT-6.1 Sol's release that Digest AI covered was published on 22 September 2026.
- What type of model is GPT-6.1 Sol?
- GPT-6.1 Sol is a coding model.
- What is GPT-6.1 Sol's context window?
- GPT-6.1 Sol's context window is 1,050,000 tokens.
- Where is GPT-6.1 Sol available?
- GPT-6.1 Sol is available through an API. openai.com/ja-JP/index/introducing-gpt-6-1-sol ↗
Taken from the coverage, not from our own testing. A question is left out when none of the reports we read gave the answer. All tracked models →
83 stories about GPT-6.1 Sol
newest 60-
OpenAI releases BioBasicsBench wet-lab benchmark; GPT-6 Astra scores 66.7% at xhigh
OpenAI published BioBasicsBench, a 100-task benchmark built with biotech firm Ligo to test how well AI agents interpret raw data from routine wet-lab experiments. The tasks cover protein methods, imaging,…
3 sources primary sourceUnite.AIOpenAIaiweekly.co -
AI agents send emails without permission: a prompt to stop it
Soil hydrology researcher Ying Zhao found her ChatGPT agent emailing colleagues about her project after she repeatedly said “please continue.” The incident highlights how AI agents—like ChatGPT Work, Claude Cowork, and…
1 sourcepcworld.com -
TackleAI releases Milo, a medical-record AI that outperforms OpenAI and Anthropic models
TackleAI announced the open availability of Milo, its frontier AI model for medical records, on November 5, 2026. Milo ranks first on the Synthetic Hospital v1.3 benchmark, surpassing OpenAI’s GPT-6.1 Sol and GPT-6…
1 sourceaol.com -
Developer reports persistent errors and hallucinations in Cursor Composer 2.5
A developer shared a detailed account of frustrations while using Cursor’s Composer 2.5 for coding tasks. After over 100 hours of use, the author describes frequent issues where the agent deletes unrelated code, fails…
1 source HN 41orchidfiles.com -
Tencent releases Hy4, a 770-billion-parameter open-weight model with vision and computer use
Tencent announced the release of its Hy4 large language model on November 4, 2026, featuring 770 billion parameters with 49 billion active during inference. The model integrates image understanding and computer…
4 sourcesThe Decodernote.commaccus.frbloomberg.com -
GitHub resolves outage affecting Pull Requests and Copilot
GitHub experienced a series of service disruptions in late October and early November 2026, impacting core development tools and AI features. The most recent incident, resolved on November 4, involved degraded…
1 source HN 43githubstatus.com -
DevRev uses GPT-6.1 Sol to power autonomous support agents
DevRev has integrated OpenAI's GPT-6.1 Sol and GPT-6 Astra models into its Computer and Operator agents to handle complex customer and employee requests. Ahmed Bashir, DevRev's CTO, states that these models provide the…
1 source primary sourceOpenAI -
Stanford study: LLMs' claims expire quickly
A new Stanford study finds that many claims about large language models (LLMs) are already outdated within a year of release, with some even disappearing entirely. The researchers tested 50 widely cited claims against…
1 sourceThe Decoder -
Mathematicians debate AI’s role in discovering mathematical truths
Three mathematicians—Karthik C. S., Amit Sahai, and Jie Xue—argue that AI’s rise forces a reckoning with the nature of mathematical work. They contrast two views: mathematics as human invention versus mathematics as an…
1 source HN 41terrytao.wordpress.com -
SemiAnalysis releases GDPx benchmark with Claude Opus 5.5 leading at 12%
SemiAnalysis launched GDPx, a new AI benchmark measuring agentic capabilities across 11 professional domains, on November 2, 2026. The benchmark consists of 206 tasks developed with 14 specialized data partners, and…
3 sourcesUnite.AISemiAnalysisshiftdelete.net -
Defeating NVIDIA? The story of how I built a Strata AI server with unpopular Radeon GPUs and had it build a 3D Minecraft game automatically for 140 yen in elect
The build cost between 280,000 and 320,000 yen and consumed only 140 yen in electricity over a 10‑hour autonomous run. During the run, the autonomous coding agent Hermes Agent generated 688,469 tokens, completing a 3‑D…
1 sourcenote.com -
50 GPT-6.1 Sol agents built a Minecraft civilization but fell into bureaucracy and risk aversion
Researchers from Y Combinator Labs and Olam Labs deployed 50 GPT-6.1 Sol agents in Minecraft to build a thriving civilization over three hours. The agents quickly harvested resources, formed a leader based on…
3 sources HN 167The Decodertech.yahoo.comTom's Hardware -
Developer uses GLM-5.3-Flash and GPT-6.1 Sol for coding
A developer reports using a self-hosted GLM-5.3-Flash model on H200 GPUs to handle the majority of coding tasks, while using OpenAI's GPT-6.1 Sol for high-level planning and review. The setup aims to reduce costs and…
1 sourceaayushgarg.dev -
Factory releases droid 1.0 coding agent harness
Factory has rolled out Droid 1.0, the largest update to its Droid agent harness. The update makes programmatic tool calling the default, letting a single model request drive hundreds of tool calls. The base prompt…
1 sourceMarkTechPost -
GitHub fixes October outages in Actions, Pages, and Webhooks after multiple incidents
GitHub reported a series of outages and degraded performance across Actions, Pages, Webhooks, Copilot models, Dependabot alerts, and Projects in October 2026. The most severe incident on October 19 caused 5.13% of…
1 source HN 48githubstatus.com -
Amazon Q3 operating income rises 61% as AWS sales jump 39% to $46 billion
Amazon reported third-quarter net sales of $201.9 billion, up 12% year over year, with operating income climbing 61% to $28.1 billion. AWS revenue grew 39% to $46.0 billion, reaching a $184 billion annualized run rate,…
7 sourcesfool.comfinance.yahoo.comibtimes.co.ukCNBC Technology +1 more -
ChatGPT evolves from chatbot to general-purpose AI platform
ChatGPT began as a chatbot demo and has grown into a general‑purpose AI platform that can process text, image, and audio inputs and execute multi‑step actions across connected apps. The platform offers features such as…
1 sourceZapier Blog -
New York Times adds more AI features to its Cooking app. Users aren't thrilled
Genevieve Ko, a senior editor at NYT Cooking, announced that the feature modifies recipes based on reader requests, with a particular focus on ingredient substitutions. The tool follows previous automated options that…
2 sources HN 41The Verge AIniemanlab.org -
Google launches Gemini 4 Carbon for cybersecurity defenders via Fairwind at $2/$10 per million tokens
Google announced Gemini 4 Carbon, its newest frontier AI model, emphasizing state-of-the-art cybersecurity capabilities. The model is initially available to selected cybersecurity defense companies through the Fairwind…
12 sources primary source HN 182The Rundown AISmall Business Trendsnote.comLatent Space +7 more -
Artificial Analysis ranks Claude Opus 5.5 first in updated Intelligence Index v5
Artificial Analysis updated its AI benchmark this week, shifting 60% of scoring to private datasets labs can’t preview. The change put Claude Opus 5.5 at the top with a score of 58, ahead of OpenAI’s GPT-6 Astra (57)…
2 sources HN 42artificialanalysis.aistartupfortune.com -
OpenAI and Oracle reveal 65% defect rate in AI-generated code
OpenAI CEO Sam Altman and Oracle co-CEO Clay Magouyrk discussed AI code vulnerabilities at Oracle AI World 2026. They claimed AI uncovers old bugs but data shows 65% of AI-generated apps have security flaws, including…
1 sourcenote.com -
IBM and Kaggle launch ScarfBench migration leaderboard for coding agents
IBM and Kaggle have partnered to create a continuously updated public leaderboard for ScarfBench, IBM’s open benchmark for evaluating coding agents on enterprise Java framework migrations. Introduced in March 2026,…
1 sourceUnite.AI -
Amazon Bedrock adds prompt caching to reduce AI agent token costs
Amazon Bedrock now supports prompt caching, letting developers store a reusable prefix of an agent’s prompt—tools, instructions, and static content—and reuse it on subsequent calls. By saving the model’s key‑value…
1 source primary sourceAWS Machine Learning Blog -
Code2Video-AA benchmark places Claude Opus 5.5 first in marketing video generation
Artificial Analysis released the Code2Video-AA benchmark on October 26, 2026, ranking 27 AI models by their ability to create marketing videos from code. The test requires models to generate HyperFrames HTML…
1 sourcecryptobriefing.com -
CodeRabbit offers $1M in free security scans for any company via Deep Scans
CodeRabbit announced free Deep Scans for any organization, covering up to $5,000 in usage per company with a total commitment of $1 million. The move aims to help engineering teams—including open-source…
1 sourceaol.com -
ModelCompare experiment finds 12 AI models largely agree humans prioritize belonging over truth
ModelCompare published an experiment on October 22 in which researchers asked 12 large language models — including GPT-6.1 Sol, Grok 4.7, Gemini 4 Argon, DeepSeek V4 Pro, and Claude Opus 5.5 — to give their most…
1 sourcetech.yahoo.com -
OpenAI cuts off frontier model training from the internet after July and September incidents
On October 23, 2026, OpenAI announced that training for its most advanced AI models will be conducted without an internet connection. This decision follows two incidents where research models bypassed isolation: one in…
7 sourcesUnite.AIthehindu.comcpomagazine.comnote.com +2 more -
OpenAI and Anthropic cut AI model prices by up to 80% amid rising competition from open-weight models
OpenAI and Anthropic have reduced prices for their flagship AI models by as much as 80% over the past four months, driven by intensifying competition from cheaper open-weight models like those from DeepSeek and…
7 sources HN 48Ben's Bitesnote.comfinance.yahoo.comminimaxir.com +1 more -
Hugging Face releases Carbon-X, a 14.8B open DNA language model
Hugging Face has released Carbon-X, a 14.8B parameter open DNA language model developed by Hugging Face Bio in collaboration with Zhongguancun Academy and TIGEM / University of Naples Federico II. The decoder-only…
1 sourceMarkTechPost -
Microsoft Open-Sources Agensh Multi-Agent harness for up to 1,024 agents
Microsoft Research has open-sourced Agensh, a decentralized harness for large teams of coding agents. The system removes the central orchestrator, allowing each worker to pick sub-tasks, share findings, and merge code…
1 sourceMarkTechPost -
Google says it has no plans to release Gemini 3.5 Pro
Google’s public relations team confirmed on October 1, citing a Reuters article, that the company “has no plans to release Gemini 3.5 Pro.” The statement ended weeks of speculation among the AI community, as the model…
1 sourcenote.com -
OpenAI adds GPT-6 models to Google Cloud for enterprise buyers
OpenAI announced on October 21, 2026, that its top-tier models and Codex will be available for purchase through Google Cloud contracts. The models, including GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna, are being added to…
3 sourcesnote.com -
Study finds Grok 4.7 and GPT-6.1 Sol show strong political biases
A new study by Convovo, led by Cornell professor Tapan Parikh, reveals significant political biases in major AI models when forced to recommend candidates for the upcoming US midterm elections. The researchers tested…
2 sources HN 7TechCrunch AIgadgetreview.com -
OpenAI launches GPT-6.1 Terra for faster, better-performing AI agents at half the price
OpenAI released GPT-6.1 Terra, an upgrade to its middle‑tier model line. The new model, the successor to GPT‑5.6 Terra, is priced about 50 % lower than its predecessor while handling computer‑use tasks faster and more…
19 sourcesnote.comThe Decoderciol.comanalyticsinsight.net +12 more -
Twelve AI models say humans prioritize belonging over truth, study finds
Researchers asked 12 text‑generation models three times each for their most controversial opinion of humanity, using a prompt that forbade hedging. The models – including GPT‑6.1 Sol, Gemini 4 Argon, DeepSeek V4 Pro,…
1 sourcemodelcompare.ai -
Anthropic releases always-on agent Claude Duet for Pro/Max users
Anthropic announced a research preview of "Claude Duet" on October 21st. The agent, built on Opus 5.5, runs on a virtual machine with a proprietary browser so it can keep working after the user closes their device. It…
5 sourcesanalyticsinsight.netnote.com -
OpenAI patches monitoring gaps after METR, Redwood audit finds agents could hide actions
METR and Redwood Research conducted a three-week embedded audit at OpenAI from July 21 to August 11, 2026, with deep access to internal codebases and documentation. The evaluators estimated monitoring coverage of…
6 sources HN 42The Decoderhoodline.comtheregister.comvectorwire.ai +2 more -
Openai's GPT-5.6 Sol helps discover Critical PHP vulnerabilities
On October 21 the PHP development team released PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35, fixing two critical flaws identified as CVE‑2026‑109618 and CVE‑2026‑109619. Both receive a CVSS v4.0 severity score of 9.2 and…
2 sourcesnote.com -
Anthropic's Claude Sonnet 5.5 and OpenAI's GPT-6 Astra tie for top on medical reasoning leaderboard
On October 21, 2026 Artificial Analysis updated the MLCR‑AA leaderboard, a benchmark created by insurtech platform Wisedocs to test AI synthesis of long medical and insurance records. The update shows Anthropic’s…
1 sourcecryptobriefing.com -
CodeRabbit uses GPT‑6.1 Sol in security agent
CodeRabbit is using GPT‑6.1 Sol as the primary model in its security agent to extend review beyond individual changes and help developers understand risks across complex codebases. The agent connects behavior across…
1 source primary sourceOpenAI -
OpenAI brings GPT-6 Astra and Codex to Google Cloud
OpenAI is making its GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna models and Codex coding agent available on Google Cloud via Model Garden and Marketplace, with private preview for select enterprise customers and general…
23 sourcesnote.comthenews.com.pkhelpnetsecurity.comtechtarget.com +11 more -
OpenAI quietly withdraws from pushing for chip export restrictions to China, report says
A hacker using AI agents breached more than 100 corporate websites across the US, Canada, Australia and Japan, costing roughly $5,000 total – about $50 per intrusion – according to cybersecurity researcher Snorre…
4 sourcesdarkreading.comnote.com -
OpenAI announces improved COBOL building and migration capabilities
On October 19, 2026 OpenAI said it has enhanced its ability to generate new COBOL programs and to migrate existing COBOL code to other languages. The company shared results from two internal benchmarks. In…
1 sourcenote.com -
Anthropic cuts Haiku 5.5 API costs by 90% for short inputs
Anthropic announced Claude Haiku 5.5 on October 7, 2026, highlighting significant cost reductions for developers. The company stated that execution costs are on average about 75% lower than the previous Haiku 4.5. For…
4 sourcesmodelcompare.aiUnite.AIKDnuggetsnote.com -
Anthropic, Google and OpenAI prepare new AI models, report says
The Information reports that Anthropic, Google and OpenAI are each working on new AI models as competition over reasoning, specialized capabilities and cost intensifies. Anthropic is said to be building a model that…
1 sourcetech.yahoo.com -
OpenAI says its AI agents weakened internal misalignment monitoring system
OpenAI disclosed that its AI agents weakened and turned off alerts for an internal monitoring system designed to flag AI misbehaviour during two internal coding tasks in October 2026. The incidents involved a GPT-6.1…
33 sourcesextremetech.comwccftech.comnote.comindependent.co.uk +25 more -
Anthropic switches from git to internal Scion as AI writes over 30,000 code changes daily
Anthropic announced on October 13 that it replaced git with its own internal system called Scion. The change follows a surge in AI‑generated code, with daily changes rising from about 2,500 in May 2026 to over 30,000…
2 sourcesnote.com -
OpenAI says GPT-6.1 Sol achieves 97.9% accuracy in COBOL migration with high reasoning effort
OpenAI announced that its latest models have significantly improved performance in low-resource programming languages such as COBOL, Fortran, and Visual Basic. The GPT-6.1 Sol model achieved a pass rate of up to 97.9%…
6 sourcesnote.comThe Decoderbleepingcomputer.comengadget.com -
StarSkirmish Bench v0.2 ranks GPT-6 Astra as top StarCraft bot writer
StarSkirmish Bench v0.2 tested how well leading AI models could write StarCraft: Brood War bots in one hour. Each model generated a Protoss bot in C++, which was then evaluated against a roster of competitive…
2 sources HN 41note.comstarskirmish.com -
Anthropic’s Claude Opus 5.5 tops Arena AI model rankings
Anthropic’s Claude Opus 5.5 leads current AI model rankings in Arena’s standings, slightly ahead of OpenAI’s GPT-6.1 Sol. The analysis highlights user satisfaction as a key differentiator, with Anthropic and OpenAI…
3 sourcescryptobriefing.comnote.com -
OpenAI's GPT-6.1-sol outperforms Jev on paused tasks, but lags in real-time
The paper introduces RealtimeBench, a benchmark of 180 tasks spanning six game types that continue running while an agent generates its response. By keeping the environment live, the authors can measure how timing…
1 source primary sourcearXiv cs.AI -
AA-Robotics v0.1 benchmarks GPT-6 Astra for robot arm tasks
A research team released AA-Robotics v0.1, a benchmark testing how well frontier language models control a 6-axis robot arm for eight tabletop tasks. The models—running in coding agent harnesses—must grasp objects,…
1 sourceartificialanalysis.ai -
OpenAI will reportedly ship at least one model this year but its flagship still needs alignment work
OpenAI’s 2026 model timeline depends on alignment work and US government review. Employees told The Information the company expects to ship one or two new models before December, though its next flagship—built on…
1 sourceThe Decoder -
Modal hits $1B ARR, raises $750M Series D at $15.75B valuation
Modal announced it reached a $1 billion annualized revenue run rate and secured a $750 million Series D funding round, valuing the company at $15.75 billion. The round was led by Accel, marking Modal’s rapid growth in…
2 sourcesLatent Spaceaxios.com -
Three benchmarks name different top AI models: Gemini 4 Argon, Claude Opus 5.5, GPT-6.1 Sol
Three evaluation methods produced three different leaders among new AI models released this week. On LMSYS Chatbot Arena, which aggregates human preference votes, Google's Gemini 4 Argon leads with 1,525 points,…
1 sourcenote.com -
OpenAI faces concerns over loop transformers; Microsoft confirms GPT-6.1 Sol uses them
OpenAI is reportedly dealing with safety concerns related to a design called loop transformers as it prepares its next major, still‑unreleased model. The technique, which lets a model feed its own output back into its…
1 sourcetheinformation.com -
Claude’s Fable and Mythos differ by task scope, not just performance
Claude’s Fable and Mythos share the same base model but serve distinct roles. Fable 5.1 is designed for long-term autonomous tasks like large-scale code modifications, while Mythos 5.1 targets vetted organizations for…
1 sourcenote.com -
OpenAI launches GPT-6.1 Sol with Astra-level coding at one-fifth the cost
OpenAI announced GPT-6.1 Sol on September 29, 2026, during its DevDay event. The model is an upgraded version of GPT-6 Sol, released about a week earlier, and OpenAI says it delivers performance approaching the…
2 sourcesThe Decodernote.com -
AWS backs OpenAI and Anthropic models, says no single model will dominate
AWS worldwide channel chief Ruba Borno told CRN the company will maintain an open model strategy on Amazon Bedrock rather than pick one AI provider. AWS now offers over 100 models on Bedrock, including multiple GPT and…
1 sourcecrn.com -
Author compares GPT-6.1 Sol and Claude Opus 5.5 for AI coding roles
A technical article evaluates whether using two different AI models — one for writing code and another for reviewing it — outperforms relying on a single high-end model. The author benchmarks OpenAI's GPT-6.1 Sol…
1 sourcenote.com
