🔍 Read the full analysis: GPT-6.1 Terra May Be Less Brilliant But Easier On Your Budget on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI released GPT-6.1 Terra on October 26, 2026, pricing API use at $1 per million input tokens and $5 per million output tokens—half the listed prices of its predecessor, GPT-5.6 Terra. Artificial Analysis found that Terra’s low and medium reasoning settings offer strong cost efficiency, while its maximum setting scores below several larger models and costs more than low effort per task.
OpenAI released GPT-6.1 Terra on October 26, setting API prices at $1 per million input tokens and $5 per million output tokens, half the listed rates for GPT-5.6 Terra. Independent benchmarking by Artificial Analysis found that Terra’s low and medium reasoning settings can deliver comparatively low costs per completed task, but its maximum-effort score remains below several larger models.
Terra is available through the API as gpt-6.1-terra and to Plus, Pro, Enterprise and Edu subscribers in ChatGPT Work and Codex, according to the launch report. OpenAI describes it as a fast, efficient option for communication and computer-based work, including processing invoices. The company says its subscription allowance can cover substantially more Terra tasks than use of Sol or Astra, though it did not provide a universal task-count figure in the supplied material.
Artificial Analysis gave Terra a 48.1 Intelligence Index score at maximum effort, with an average cost of $0.40 per task across its evaluation tasks. On the same measure, GPT-6.1 Sol scored 51.8 at $0.72 per task, while Claude Opus 5.5 scored 57.6 at $5.98. The index combines ten evaluations; these figures are the firm’s reported averages, not a price guarantee for every user’s workload.
Terra has six effort settings. The supplied results show a range from 33.5 points and $0.107 per task in non-reasoning mode to 48.1 points and $0.40 at maximum effort. Low effort was reported at 36.2 points and $0.055 per task. Artificial Analysis found that low and medium settings fell on its cost-efficiency frontier, while settings above them did not. The source material does not include every row of the six-setting table, so the exact medium, high and extra-high figures cannot be stated here.
Infographic 01 · The effort curve
One model, six settings, a 7-fold cost range.
GPT-6.1 Terra at each effort level, as measured by Artificial Analysis. Low and medium sit on its cost-efficiency frontier: no tested model is both cheaper and better.
Score · Intelligence Index
independent higher is better
Cost per finished task
note the last step shorter is cheaper
All six settings
| Effort | Index | Cost / task | Output tokens / task | Hallucination rate |
|---|---|---|---|---|
| Non-reasoning | Index 33.5 | Cost $0.107 | Tokens 4,737 | Hallucination 72% |
| Low | Index 36.2 | Cost $0.055 | Tokens 2,681 | Hallucination 46% |
| Medium | Index 41.3 | Cost $0.095 | Tokens 5,835 | Hallucination 44% |
| High | Index 43.9 | Cost $0.134 | Tokens 9,027 | Hallucination 43% |
| Xhigh | Index 45.4 | Cost $0.194 | Tokens 14,551 | Hallucination 42% |
| Max | Index 48.1 | Cost $0.403 | Tokens 35,096 | Hallucination 38% |
Hallucination rate in AA-Omniscience: how often the model answers anyway when it does not know. Lower is better.
The first steps up are cheap. The last one doubles the bill.Assessment: Thorsten Meyer
Sources: Artificial Analysis model pages (retrieved 26 Oct 2026); Artificial Analysis launch article (26 Oct 2026). Cost per task is the average across the index’s evaluation tasks.
Powered by Thorsten Meyer AI
Task Costs Change the Comparison
Terra’s launch highlights a difference between token pricing and the cost of completing work. A model with cheaper tokens can still use more of them, while another may finish a task with fewer tokens. Artificial Analysis reported that GPT-6 Luna’s listed price per token is ten times lower than Terra’s, yet Terra at low effort costs less per task in its evaluation. That result applies to the benchmark’s tasks and measurement method; it does not establish which model is cheapest for every workflow.
For businesses processing large volumes of routine work, a low cost per completed task could matter more than a headline rate per million tokens. The trade-off is capability: Terra’s maximum-effort score is below Sol’s and several competing models, and the source says Xiaomi’s open-weight MiMo-V2.6-Pro scored 46.3 for $0.13 per task—higher and cheaper than Terra’s high and extra-high settings. Teams may need to test their own workload and account for quality, review and failure costs before choosing a model.
Products Worth Considering
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How Terra Fits OpenAI’s Lineup
OpenAI positions Terra between its smaller GPT-6 Luna and larger Sol and Astra models. The launch report says Terra’s API prices are also half those of GPT-6.1 Sol, at $2 per million input tokens and $10 per million output tokens. This is a pricing and product-line distinction, not evidence that Terra replaces the higher-capability models.
OpenAI’s own launch-page figures make separate claims about task types. On Agents’ Last Exam, the company reports Terra at medium effort scored 57.6% and cost $0.29 per task, compared with GPT-5.6 Terra at 42.6% and $0.84. For maximum-effort computer use, OpenAI reports Terra at 60.0% and $1.01 per task, compared with Claude Opus 5.5 at 63.2% and $8.82. These are company-reported benchmark results, distinct from Artificial Analysis’s independent index.
OpenAI also reports improved results over GPT-5.6 Terra on an internal set of prompts designed to induce factual errors, and a 50.3% score on Terminal-Bench 4.0 at maximum effort, compared with the predecessor’s published 21.5%. The supplied report says OpenAI still recommends Sol and Astra for the most complex coding tasks. Its speed claim—that Terra completes tasks more than three times faster than those models—was not independently measured by Artificial Analysis at the time covered by the report.
Workload Results Still Vary
The reported cost per task is an average across Artificial Analysis’s evaluation set. Actual API bills will depend on task mix, token use, reasoning setting and cached input; the listed cached-input price is $0.05 per million tokens. The supplied material does not provide the complete results for all six effort settings, making it impossible to compare every point on Terra’s cost-and-score curve.
Some performance claims are from OpenAI’s own tests, while the Intelligence Index figures come from an external evaluator. They use different benchmarks and should not be treated as directly interchangeable. Artificial Analysis had not published speed measurements for Terra when the launch report checked. The report also does not give enough detail to independently evaluate OpenAI’s internal factual-error test or its cybersecurity classification.
Testing Terra in Real Workflows
The next practical test is how Terra performs on customer workloads, where completion rates, correction time and review costs can change the economics suggested by benchmark averages. Organizations considering the model can compare low and medium effort with their existing tools, then reserve higher effort or larger models for tasks where the extra capability justifies the added cost.
Artificial Analysis may add further measurements, including speed data, while users and companies gather results from ChatGPT Work, Codex and the API. Until those comparisons are available, Terra’s price and benchmark results are confirmed in the launch report, but its value for any specific workflow remains a matter for testing.
Key Questions
How much does GPT-6.1 Terra cost?
OpenAI’s listed API price is $1 per million input tokens and $5 per million output tokens. Cached input is listed at $0.05 per million tokens. Actual task costs vary with token use and the selected effort setting.
Is Terra OpenAI’s highest-scoring model?
No. Artificial Analysis scored Terra at 48.1 at maximum effort, below GPT-6.1 Sol at 51.8 and GPT-6 Astra at 52.7 in the supplied comparison. The source report also identifies several non-OpenAI models with higher scores.
Which Terra effort setting is cheapest?
Among the settings with figures provided in the source material, low effort cost $0.055 per task, compared with $0.107 for non-reasoning and $0.40 for maximum effort. The full six-setting table was not supplied, so those figures do not establish the cost of every setting.
Where can people use GPT-6.1 Terra?
The launch report says Terra is available through the API as gpt-6.1-terra and in ChatGPT Work and Codex for Plus, Pro, Enterprise and Edu users.
Does a lower price mean Terra is best for every task?
No. Artificial Analysis found strong cost efficiency at low and medium effort, but Terra scored below several larger models at maximum effort. OpenAI says Sol and Astra remain better choices for its most complex coding tasks.
Source: ThorstenMeyerAI.com
