OpenAI has launched GPT-6.1 Terra, a quick mannequin for visible reasoning, laptop use and writing. It replaces GPT-5.6 Terra and is reside within the API as gpt-6.1-terra. Customary pricing is $1 per million input tokens and $5 per million output tokens. For groups automating screens, types and paperwork at quantity, GPT-6.1 Terra modifications the associated fee math.
TL;DR
- Dimension: 1.05M-token context, 128K max output.
- Runs on: OpenAI-hosted solely (API, ChatGPT Work, Codex). Closed weights.
- Efficiency: Close to GPT-6.1 Sol and GPT-6 Astra on imaginative and prescient and laptop use, at decrease price.
- Finest: 85.5% on MMMU-Professional (Imaginative and prescient) at max effort, 1.2 factors behind GPT-6 Astra.
- Worst: 50.3% on Terminal-Bench 4.0, behind GPT-6.1 Sol (55.9%) and Claude Sonnet 5.5 (70.6%).
What’s GPT-6.1 Terra?
GPT-6.1 Terra is an OpenAI reasoning mannequin constructed for high-volume visible and computer-use work. It accepts text and image input and returns text. The context window is 1,050,000 tokens, with as much as 128,000 output tokens. Its data cutoff is June 1, 2026.
Reasoning effort runs from none to max, with medium because the default. Instrument calling requires the Responses API. Supported instruments embody laptop use, hosted shell, code interpreter, net search, file search and MCP. OpenAI’s listed use instances embody matching product images to catalog entries and decoding technical drawings.
How does GPT-6.1 Terra carry out on laptop use?
OpenAI’s principal influential check is Agents’ Last Exam v1, a set {of professional} laptop duties. One job turns architectural drawings right into a 3D mannequin.
At medium effort, Terra scores 57.6% at $0.29 per job. GPT-5.6 Terra scored 42.6% at $0.84 on the identical setting. GPT-6 Astra additionally scores 57.6% at medium, however prices $4.82 per job.
At max effort, Terra reaches 60.0% at $1.01 per job. Claude Opus 5.5 scores 63.2% at $8.82 per job on the identical chart. OpenAI says Terra completes duties over 3 occasions quicker than GPT-6 Astra and GPT-6.1 Sol.
On the OSWorld 2.1 offline set, Terra scores 66.7% at medium and 74.2% at max. GPT-5.6 Terra peaked at 57.1%. GPT-6.1 Sol reaches 78.0%, GPT-6 Astra 80.9% and Claude Sonnet 5.5 83.9%.
How sturdy is its visible reasoning?
On MMMU-Professional (Imaginative and prescient), Terra scores 85.5% at max effort. GPT-6.1 Sol scores 86.3% and GPT-6 Astra 86.7%. GPT-5.6 Terra topped out at 81.7%. Terra’s max-effort run prices $0.012 per job, in opposition to $0.065 for Astra.
window.addEventListener(“message”,operate(e){if(e.information&&e.information.mtpTerraH){var f=doc.getElementById(“mtp-t61-frame”);if(f){f.type.top=e.information.mtpTerraH+”px”;}}});
Did factuality and coding enhance?
On OpenAI’s inside set of error-inducing prompts, the low-effort error price fell from 53.6% to 16.7%. At max effort, it fell 18 factors to 4.3%. The set is adversarial by design, so manufacturing error charges will differ.
On Terminal-Bench 4.0, Terra solves 50.3% at max effort, up from 21.5% for GPT-5.6 Terra. It nonetheless trails GPT-6.1 Sol (55.9%), GPT-6 Astra (58.2%) and Claude Sonnet 5.5 (70.6%). OpenAI itself recommends Sol or Astra for probably the most complicated coding duties.
On Terminal-Bench-Science 0.1, Terra scores 58.3% at $4.15 per job. That beats GPT-6.1 Sol’s 57.0% at $5.47. Claude Opus 5.5 with fallbacks scores 63.3% at $23.21.
How a lot does GPT-6.1 Terra price, and the place is it out there?
Standard API pricing is $1 enter, $0.05 cached enter and $5 output per million tokens. Batch and Flex are 50% cheaper. Quick mode prices 2x Customary. Prompts above 272K enter tokens invoice at 2x enter and 1.5x output.
That’s half of GPT-6.1 Sol’s $2/$10 and one-tenth of GPT-6 Astra’s $10/$50. Claude Opus 5.5 lists at $4/$20.
In ChatGPT, Terra is reside for Plus, Professional, Enterprise and Edu customers in ChatGPT Work and Codex.
GPT-6.1 Terra vs rivals
| Function | GPT-6.1 Terra | GPT-6.1 Sol | GPT-6 Astra | Claude Opus 5.5 | Claude Sonnet 5.5 |
|---|---|---|---|---|---|
| Parameters | Not disclosed | Not disclosed | Not disclosed | Not disclosed | Not disclosed |
| Context window | 1.05M | 1.05M | 1.05M | 1M | 1M |
| Max output | 128K | 128K | 128K | 128K | 128K |
| Data cutoff | Jun 1, 2026 | Apr 30, 2026 | Apr 30, 2026 | Jun 2026 | Jun 2026 |
| API value (in/out per 1M) | $1 / $5 | $2 / $10 | $10 / $50 | $4 / $20 | $2 / $10 |
| Weights | Closed | Closed | Closed | Closed | Closed |
| Brokers’ Final Examination v1 (max) | 60.0% ($1.01/job) | 58.6% ($2.00/job) | 59.3% ($7.24/job) | 63.2% ($8.82/job) | Not proven |
| MMMU-Professional Imaginative and prescient (max) | 85.5% | 86.3% | 86.7% | Not proven | Not proven |
| OSWorld 2.1 offline (max) | 74.2% | 78.0% | 80.9% | Not proven | 83.9% |
| Terminal-Bench 4.0 (max) | 50.3% | 55.9% | 58.2% | Not proven | 70.6% |
| Terminal-Bench-Science 0.1 (max) | 58.3% | 57.0% | 68.1% | 63.3% (with fallbacks) | Not proven |
Benchmark scores and per-task prices are learn from OpenAI’s launch charts and are self-reported. Specs and costs come from every vendor’s mannequin web page. Claude costs: Anthropic pricing.
Key Takeaways
- GPT-6.1 Terra prices $1/$5 per million tokens, half of GPT-6.1 Sol.
- It matches GPT-6 Astra’s 57.6% on Brokers’ Final Examination at $0.29 per job.
- MMMU-Professional hits 85.5%, inside 1.2 factors of Astra.
- Coding nonetheless lags: 50.3% on Terminal-Bench 4.0 vs Sonnet 5.5’s 70.6%.
- All scores are OpenAI-reported; parameters and {hardware} are undisclosed.
Try the OpenAI announcement, the API model page and the system card addendum. Additionally, be happy to comply with us on Twitter and don’t neglect to hitch our 150k+ML SubReddit and Subscribe to our Newsletter. Wait! are you on telegram? now you can join us on telegram as well.
Must accomplice with us for selling your GitHub Repo OR Hugging Face Web page OR Product Launch OR Webinar and many others.? Connect with us
The publish OpenAI Launches GPT-6.1 Terra, Halving GPT-6.1 Sol’s API Price for Agent Work appeared first on MarkTechPost.

