OpenAI Releases GPT-6.1 Terra With 50% API Price Cut and 3x Faster Computer Use
OpenAI released GPT-6.1 Terra, a fast, efficient model focused on computer operation, visual reasoning, and writing, rolling out in ChatGPT Work and Codex to Plus, Pro, Enterprise, and Edu users and via API. OpenAI claims computer-operation tasks complete more than three times faster than GPT-6 Astra and GPT-6.1 Sol, with API prices cut 50% to $1 per million input tokens and $5 per million output tokens. Independent evaluations report mixed results: Artificial Analysis calls it Pareto-optimal for intelligence versus cost, while ToneBench ranks it #23 of 84 models, behind GPT-5.6 Terra.
Reference imageEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
GPT-6.1 Terra and GPT-6 Astra Tie at 57.6% as Costs Diverge 16.6x
An X post reports that OpenAI's GPT-6.1 Terra and GPT-6 Astra both score 57.6% on OpenAI's Agents' Last Exam at medium effort, while their per-task costs differ sharply: Terra at $0.29 a task versus Astra at $4.82, a 16.6x gap. The post also cites OpenAI's own page stating that for the most complex coding tasks, GPT-6.1 Sol and GPT-6 Astra "remain the better choice." The item presents a benchmark parity result alongside a large cost differential and OpenAI's stated guidance favoring Sol and Astra for the hardest coding work.
Read sourceGPT-6.1 Terra ranks #23 on ToneBench, trails GPT-5.6 Terra at #15
A ToneBench evaluation of OpenAI's new GPT-6.1 Terra found it ranks #23 of 84 models, behind GPT-5.6 Terra at #15, despite costing half as much. Only GPT-5.6 on ultra clearly beats the new model; the other three configurations overlap in Elo range and are effectively tied, with 6.1 on low matching 5.6 on medium for less than half the cost. The effort setting reportedly makes little difference: on low, 6.1 Terra uses zero thinking tokens per draft, while on max it uses about 9,500 tokens and costs roughly 3x as much for the same score. Only one of four LLM judges, OpenAI's GPT-6.1 Sol, clearly prefers 6.1 Terra. The model ranks #10 of 141 configs for avoiding slop but falls short on tone and voice (#64) and YouTube practices (#68). The author suggests using it on low for cheap bulk first drafts, while GPT-6/6.1 Sol on medium beats every Terra setting at about twice the cost of 6.1 on low. Cheaper open-weight alternatives GLM-5.3 Flash and DeepSeek V4.1 Flash 1017 write better for less but are slower.
Read sourceOpenAI Releases GPT-6.1 Terra, Claiming 3x Faster Professional Task Completion
OpenAI has released a new model, GPT-6.1 Terra, according to a breaking post. The company claims Terra completes professional computer tasks more than 3x faster than both GPT-6 Astra and GPT-6.1 Sol, while operating at substantially lower costs. The announcement frames the release around speed and cost efficiency for professional computer-task workloads, positioning Terra against two prior OpenAI models. No independent benchmarks, pricing figures, availability details, or technical specifications were provided in the source item, and the performance and cost claims are attributed solely to OpenAI.
Read sourceShow 9 older updatesHide older updates
OpenAI Releases GPT-6.1 Terra, Focused on Computer Use and Writing
The author announced the return of GPT-6.1 Terra, describing it as the first model that is notably better at writing, personality, and nuanced understanding of requests. According to the publisher, the model excels at computer use, able to understand screen content and perform actions across applications. The post also states that Terra offers twice the usage of GPT-6.1 Sol in both API and subscription offerings at a lower price. The publisher presented these improvements in writing and screen understanding, along with the two-times usage comparison against 6.1 Sol, to help assess suitable use cases.
Read sourceOpenAI Releases GPT-6.1 Terra With 60% Score on Agents' Last Exam
OpenAI has released GPT-6.1 Terra, which reaches roughly 60% on OpenAI's Agents' Last Exam at around $1 per task. According to the post, Opus 5.5 scores a few points higher but costs roughly 9x as much at the settings shown. OpenAI also reports that GPT-6.1 Terra completes tasks over 3x faster than GPT-6 Astra and GPT-6.1 Sol. API pricing is $1 per million input tokens and $5 per million output tokens.
Read sourceOpenAI's GPT-6.1 Terra launches on OpenRouter at $1/M input tokens
OpenAI's GPT-6.1 Terra is now available on OpenRouter, according to a post from the platform. The model reportedly matches GPT-6 Astra on the Agents' Last Exam benchmark at one-tenth the price per token, and completes tasks more than 3x faster on average than both Astra and GPT-6.1 Sol. Pricing is listed at $1 per million input tokens and $5 per million output tokens, with a 1.05M token context window. The post presents these benchmark, speed, and pricing comparisons as the model's key differentiators but does not provide independent verification or additional technical details.
Read sourceOpenAI Announces GPT-6.1 Terra, Claiming 9x Lower Cost Per Task Than Claude Opus 5.5
OpenAI officially announced GPT-6.1 Terra, according to a post on X. The announcement claims that at max effort Terra is over three times as token-efficient as Claude Opus 5.5 and costs one-fourth its price per token, resulting in roughly 9x lower cost per task. The post also states that Terra is much quicker, completing tasks over three times faster than both GPT-6 Astra and GPT-6.1 Sol. The poster characterized it as a Berlin DevDay release. These performance and pricing figures are attributed to the announcement and have not been independently verified in the source item.
Read sourceOpenAI Launches GPT-6.1 Terra with Triple Computer-Operation Speed and 50% API Price Cut
OpenAI announced GPT-6.1 Terra on October 27, a new model focused on improving computer operation, visual understanding, and programming capabilities. It can assist users with drafting proposals, interpreting technical drawings, filling out forms, and editing videos. Compared with the previous generation, the model completes computer-operation tasks more than three times faster and cuts API prices by 50%. API pricing is set at $1 per million input tokens, $0.05 per million cached input tokens, and $5 per million output tokens. Internal tests show the error rate at the low reasoning tier dropped from 53.6% to 16.7%, while the highest reasoning tier's error rate fell to 4.3%, an 18 percentage point decrease from the previous generation. In programming tests, the highest reasoning tier's score rose from 21.5% to 50.3%. The new model is available to Plus, Pro, Enterprise, and Edu users of ChatGPTWork and Codex, and is offered via API.
Read sourceOpenAI Releases GPT-6.1 Terra With Lower Pricing and Higher Intelligence Scores
OpenAI released GPT-6.1 Terra, the first Terra-class model since GPT-5.6, priced at $1/$5 per million input/output tokens versus $2/$12 previously, with cache write halved to $1.25 and cached input at $0.05 per million tokens. Artificial Analysis reports the model is Pareto-optimal for intelligence versus cost at low and medium effort; at max effort it gains 6 points on the Intelligence Index over GPT-5.6 Terra at less than a third of cost per task ($0.40 vs $1.40), and scores 4 points below GPT-6.1 Sol and 10 above GPT-6 Luna. It scores 48 on the Coding Agent Index v2.0, 6 points below GPT-6.1 Sol and 17 above GPT-6 Luna. Hallucination rate fell 50 points to 38% from 88%, alongside a 2 point accuracy gain in AA-Omniscience, and Terminal-Bench 4.0 improved about 20 points.
Read sourceOpenAI Rolls Out GPT-6.1 Terra to Plus, Pro, Enterprise and Edu Users
OpenAI announced GPT-6.1 Terra, described as a new fast and efficient model that combines sharper visual reasoning and precise computer use with clearer, more natural writing. The company said the model is built for the whole workday, citing tasks such as decoding technical drawings, drafting proposals, filling forms, and editing videos. GPT-6.1 Terra is rolling out today to Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex.
Read sourceOpenAI Releases GPT-6.1 Terra Model for Plus, Pro, Enterprise and Edu Users
OpenAI announced the release of a new model, GPT-6.1 Terra, positioned as fast and efficient, with strengthened visual reasoning, precise computer operation, and clearer, more natural writing. According to the company, the model is suited to full-workday scenarios such as interpreting technical drawings, drafting proposals, filling out forms, and editing video. OpenAI said GPT-6.1 Terra is rolling out starting today in ChatGPT Work and Codex to Plus, Pro, Enterprise, and Edu users. The announcement frames the model's capability positioning and user coverage so readers can judge whether to upgrade.
Read sourceOpenAI's GPT-6.1 Terra Model Spotted on APIs and Codex Changelog
A new OpenAI model slug, gpt-6.1-terra, has been spotted on the company's APIs and is now referenced in the latest Codex changelog, according to a post from the account testingcatalog. The post quotes the changelog as saying GPT-6.1 Terra is now available in ChatGPT Work and Codex on Plus, Pro, Business, Enterprise, and Edu plans. It describes Terra as a fast, efficient model for professional work that combines strong image understanding with precise computer use, able to reason about images and screens and then act across the browser, desktop, and apps. The quoted text also says Terra provides clear, direct writing that puts results first and explains supporting details in plain language. The poster said testing would follow soon. The information is attributed to API observations and the Codex changelog rather than a formal OpenAI announcement.