🔍 Read the full analysis: GPT-6.1 Terra: A Potentially Cheaper Way To Put AI To Work on ThorstenMeyerAI.com
Get hardware and tech essentials 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. Artificial Analysis found that the model’s low and medium reasoning settings offer a stronger cost-to-score balance than its higher settings, but Terra does not lead the benchmark and results depend on the task.
OpenAI released GPT-6.1 Terra on October 26, 2026, positioning it as a lower-cost model for routine workplace and computer-based tasks. Its API price is $1 per million input tokens and $5 per million output tokens, half the listed rates for GPT-5.6 Terra; independent testing found that the cost of a completed task varies widely with the model’s reasoning setting.
OpenAI says Terra is designed for work such as reading 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 model name gpt-6.1-terra. OpenAI says the same allowance in ChatGPT Work can cover substantially more tasks with Terra than with Sol or Astra, but the company has not provided a general allowance conversion in the supplied launch material.
Artificial Analysis, a benchmarking firm, reported an Intelligence Index score of 48.1 for Terra at maximum effort, with an average cost of $0.40 per task across its index evaluations. In the same comparison, 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 and its cost figure is an average across those tasks, not a guaranteed price for a particular customer workload.
OpenAI’s own benchmark figures also show areas where Terra improved on its predecessor. On Terminal-Bench 4.0 at maximum effort, it solved 50.3% of tasks, compared with the 21.5% published for GPT-5.6 Terra. OpenAI says Terra completes tasks more than three times faster than Astra and Sol; Artificial Analysis had not published speed measurements for Terra at the time of the source report. These results come from different evaluations and should not be treated as direct measures of performance on every real-world task.
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 Reasoning
The launch puts attention on a distinction that token prices alone can hide: the amount spent to finish a task depends on how many tokens a model uses, including reasoning and output. Artificial Analysis measured Terra’s six effort settings at costs ranging from $0.05 to $0.40 per task and scores from 33 to 48. At maximum effort, the reported cost was 7.4 times the low-effort cost for a 12-point higher index score.
That variation may matter for businesses running large volumes of repeatable work. A company evaluating invoice processing, coding assistance or other automated workflows would need to test the model on its own tasks and include error handling, review and usage patterns in its cost estimate. The benchmark supports a case for testing low or medium effort on suitable tasks; it does not establish that Terra will be cheaper than alternatives in every deployment.
The comparison also shows why a lower per-token price is not automatically a lower bill. Artificial Analysis reports that GPT-6 Luna costs $0.10 per million input and $0.50 per million output tokens, below Terra’s rates, but its listed average task cost is $0.07, compared with $0.055 for Terra at low effort. Token use differs between models, so customers should compare equivalent completed work rather than infer task costs from token rates alone.
Products Worth Considering
LG UltraGear 27G640A-B 27-Inch QHD IPS Gaming Monitor, 300Hz, 1ms, G-SYNC Compatible, FreeSync Premium, DisplayHDR 400
LG UltraGear 45GX900A-B 45-Inch Curved OLED Gaming Monitor
LG UltraGear 32GX870A-B 32-inch 4K OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Where Terra Sits in OpenAI’s Lineup
OpenAI describes Terra as a fast, efficient model for everyday communication and computer-based tasks. The launch places it between the smaller Luna model and the larger Sol and Astra models. OpenAI says that for the most complex coding tasks, Sol and Astra remain the better choice, a qualification that points to a trade-off between price, capability and task difficulty.
Artificial Analysis’s launch-day comparison did not put Terra at the top of its Intelligence Index. Claude Opus 5.5, Kimi K3.1 Pro and Gemini 4 Argon scored above Terra at the settings listed in the source. The same table includes lower-cost alternatives: for example, the open-weight MiMo-V2.6-Pro scored 46.3 at a reported average task cost of $0.13. Those figures reflect the firm’s specific test suite and listed prices, not a universal ranking for all uses.
Artificial Analysis also reported that Terra guessed on 38% of questions it could not answer in its hallucination test, compared with 88% for GPT-5.6 Terra. That result concerns the firm’s test and question set; it does not mean Terra avoids errors in general use. OpenAI separately reported lower error rates on its own set of error-inducing prompts, but those internal results are not the same evaluation.
Benchmark Results Have Limits
The supplied launch material does not include the full results for every one of Terra’s six effort settings, so the complete score-and-cost curve cannot be reconstructed from the available figures. Artificial Analysis’s scores and average costs also do not predict what a specific company will pay or how accurately Terra will perform on its own data.
Some claims remain company-reported rather than independently measured in the material. OpenAI’s speed comparison had no corresponding Artificial Analysis speed result at the time of publication, and its cybersecurity designation—Terra meets the Critical threshold under OpenAI’s Preparedness Framework—is the company’s own classification. The sources also do not provide detailed information about real-world error rates, customer uptake or total deployment costs, including human review.
Testing Terra on Real Workloads
Organizations considering Terra can compare it with other models on representative tasks, beginning with low or medium effort where the work is routine and escalating when results fall short. A useful evaluation would track cost per completed task, accuracy, speed and review time, not just the API token rate. OpenAI’s release makes Terra available through ChatGPT Work, Codex and the API; the supplied source does not identify a later launch milestone or a date for additional independent testing.
For users and companies, the remaining question is whether the reported cost and performance pattern holds outside benchmark suites. AI systems can make mistakes, and higher reasoning effort or additional human checks can raise costs. Results will depend on workload, settings and usage, so the launch figures are a starting point for evaluation rather than a promise of savings.
Key Questions
How much does GPT-6.1 Terra cost through the API?
OpenAI lists $1 per million input tokens and $5 per million output tokens. Cached input is listed at $0.05 per million tokens in the source material.
Is Terra cheaper than GPT-6.1 Sol?
Its listed API token prices are lower: $1/$5 per million input/output tokens, compared with $2/$10 for Sol. Artificial Analysis also reported a lower average cost per task at maximum effort for Terra, $0.40 versus $0.72, across its index evaluations. Actual costs depend on the task and settings.
Does Terra outperform OpenAI’s larger models?
Not on every reported measure. At maximum effort, Terra scored 48.1 on Artificial Analysis’s Intelligence Index, below Sol’s 51.8. OpenAI says Sol and Astra remain better choices for its most complex coding tasks.
Where can people use GPT-6.1 Terra?
OpenAI made it 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
