🔍 Read the full analysis: GPT-6.1 Terra Could Be A Practical Choice For Cost-Sensitive AI 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, positioning it as a lower-cost model for everyday work. Artificial Analysis measured a $0.055 average cost per task at low effort and $0.40 at max, but Terra scored below several higher-performing models and the cheapest setting was not the non-reasoning mode.
OpenAI released GPT-6.1 Terra on October 26, pricing its API at $1 per million input tokens and $5 per million output tokens—half the listed rates for GPT-5.6 Terra. Independent benchmark data from Artificial Analysis suggests Terra’s cost advantage depends on how it is used: at low effort, the firm measured an average of $0.055 per task, while max effort cost $0.40 and produced a higher score.
OpenAI describes Terra as a fast, efficient model for communication and computer-based work, including processing invoices and entering information into accounting software. It is available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, as well as through the API under the name gpt-6.1-terra. OpenAI says the same ChatGPT allowance covers substantially more tasks with Terra than with Sol or Astra.
Artificial Analysis evaluated Terra across ten tests in its Intelligence Index. At maximum effort, Terra scored 48.1 and averaged $0.40 per task. GPT-6.1 Sol scored 51.8 at $0.72 per task, while Claude Opus 5.5 scored 57.6 at $5.98. These are benchmark averages, not a guarantee of what any specific user task will cost or how well it will perform.
The tested effort settings show why token price alone can mislead. Artificial Analysis measured Terra at 36.2 points and $0.055 per task at low effort, compared with 33.5 points and $0.107 in non-reasoning mode. At max effort, the score rose to 48.1 and the cost to $0.40. The source reports six settings spanning scores from 33 to 48 and costs from $0.05 to $0.40; the low setting was cheapest among the figures provided, while non-reasoning was not.
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 Depend on Effort
For businesses running large volumes of routine AI work, the relevant expense may be the cost of completing a task, not the advertised price per token. Models can produce different amounts of text and reasoning, so a lower token rate does not by itself establish which option is cheaper for a particular workflow. Artificial Analysis’s results put Terra’s low-effort average below its own non-reasoning setting, despite the latter sounding like the lighter mode.
The trade-off is capability. Terra’s max-effort score of 48.1 is below Sol’s 51.8 and Opus 5.5’s 57.6 in the same index. Artificial Analysis also lists lower-cost alternatives, including the open-weight MiMo-V2.6-Pro, which scored 46.3 for $0.13 per task. Those comparisons make Terra a possible fit for selected workloads, not an automatic best buy across all models or tasks.
OpenAI says Terra completes tasks more than three times faster than Astra and Sol, but Artificial Analysis had not published speed measurements for Terra when the source checked. Companies weighing the launch should treat that speed statement as an OpenAI claim and test their own workloads, including accuracy, latency and total task costs.
Products Worth Considering
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
Samsung 27-Inch Odyssey QD-OLED G8 4K 240Hz Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Where Terra Fits in OpenAI’s Lineup
OpenAI places Terra between its smaller Luna model and larger Sol and Astra models. Its API rate is listed at $1/$5 per million input/output tokens, compared with $2/$12 for GPT-5.6 Terra. Cached input is priced at $0.05 per million tokens, according to the source. Artificial Analysis’s listed rates and its average cost-per-task measure use different units: the latter includes what models spend across the firm’s benchmark evaluations.
OpenAI also published task-specific results. On Agents’ Last Exam, the company says Terra at medium effort scored 57.6% for $0.29 per task, compared with GPT-5.6 Terra’s 42.6% and $0.84. On Terminal-Bench 4.0, OpenAI reports a 50.3% solve rate at max effort, against 21.5% for GPT-5.6 Terra’s published result. These are company-reported benchmark figures, distinct from Artificial Analysis’s independent index.
Artificial Analysis said Terra gained six points over GPT-5.6 Terra at less than a third of the predecessor’s average task cost: $0.40 versus $1.40 at max effort. It also reported a 48 score for Terra in its Coding Agent Index, compared with 54 for Sol and 31 for Luna. Benchmark scores reflect specific test sets and should not be read as universal measures of real-world performance.
Performance Outside the Benchmarks
The published figures do not establish how Terra will perform across every company’s workflows. Artificial Analysis’s task costs are averages over its evaluation set, and actual bills can vary with prompt length, cached input, output volume and chosen effort setting. The available source text identifies six effort levels but does not provide complete figures for every level, so a full cost-and-score curve cannot be reconstructed here.
OpenAI’s claims about speed, factual-error reductions and selected task benchmarks have not been independently verified in the material provided. OpenAI reports that its internal error-inducing prompt set showed error rates falling from 53.6% to 16.7% at low effort and from 22.3% to 4.3% at max effort compared with GPT-5.6 Terra. The test’s methodology and how those results translate to routine use remain unclear. Artificial Analysis separately reported that Terra guessed on 38% of questions it could not answer in its hallucination test, down from 88% for GPT-5.6 Terra; the source does not provide the full test details.
Artificial Analysis had not published Terra speed measurements as of the source’s October 26 check. The comparisons also do not establish whether Terra’s lower average cost will outweigh any differences in accuracy, human review needs or integration expenses for a particular customer.
Watch for Real-World Cost Data
The immediate next step for prospective users is to test Terra on representative tasks at more than one effort setting, then compare the cost of accepted, completed work rather than token rates alone. For company deployments, that means tracking output quality, retries, review time and API charges alongside task completion rates.
Further independent measurements, especially on speed and a wider range of business workflows, could clarify where Terra’s cost profile holds up. OpenAI’s launch material says Terra meets the Critical threshold in cybersecurity under its Preparedness Framework and retains safeguards developed for GPT-6 Astra; that classification is OpenAI’s own assessment. The source does not identify a later review or additional release date.
Key Questions
How much does GPT-6.1 Terra cost?
OpenAI lists API prices of $1 per million input tokens and $5 per million output tokens, with cached input at $0.05 per million tokens, according to the source. Actual task costs depend on usage and effort setting.
Which Terra effort setting was cheapest in the reported test?
Artificial Analysis measured low effort at $0.055 per task, compared with $0.107 for non-reasoning mode. The source provides a score of 36.2 for low effort and 33.5 for non-reasoning.
Is Terra more capable than OpenAI’s larger models?
Not in the Artificial Analysis Intelligence Index figures provided. At max effort, Terra scored 48.1, below GPT-6.1 Sol at 51.8. OpenAI also says Sol and Astra remain better choices for its most complex coding tasks.
Where can users access GPT-6.1 Terra?
The source says Terra is available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex, and through the API as gpt-6.1-terra.
Source: ThorstenMeyerAI.com
