[whatmodel.sh](../../README.md) / GPT-6.1 Terra # GPT-6.1 Terra OpenAI · released 2026-10 · closed weights **score 83.5** ±0.8 · #3 of 732 · from 11 benchmark results across 3 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks GPT-6.1 Terra among every model on that task or test, worst to best. 
task         vs the others         rank  counts  tests
agent tasks  ▁▁▃▆█▄▅▅▄▄●▁    #8 of 313*     26%      2
coding       ▁▁▁▃▅▇█▇▅▄●▁   #13 of 469*     11%      1
math         ▁▁▃▅█▆▄█▄▂●▁     #4 of 431     21%      3
agent tasks · #8 of 313 · 26% of its score Trails [Gemini 4 Argon](gemini-4-argon.md), [Claude Sonnet 5.5](claude-sonnet-5-5.md), [Claude Fable 5.1](claude-fable-5-1.md) here, though they rank lower overall. 
test                                    result  vs the others in its run  difficulty            counts  why
Arena Agent (October 2026 aggrega…  0.1057 IPS  ▁▁▁▃▃▂█▁▂●▁▁ beats 52/59  avg -0.379 IPS           16%  sharp
Terminal-Bench 4.0 (maker)               50.3%  ▁·▁··▁▆▆▅█●▃ beats 25/33  very hard · avg 0.4%    9.2%  sharp · at its level · small run
coding · #13 of 469 · 11% of its score Trails [Qwen3.8 Max (0902)](qwen3-8-max-0902.md), [Claude Mythos Preview (limited preview)](claude-mythos-preview.md), [GPT-6 Sol](gpt-6-sol.md) here, though they rank lower overall. 
test                       result  vs the others in its run    difficulty  counts  why
Arena WebDev (Code Arena)  1672.7  ▁▁▃▂▅█▃▆▅●▁▂ beats 117/129  avg 981        11%  sharp
math · #4 of 431 · 21% of its score Trails [GPT-6.1 Sol](gpt-6-1-sol.md) here, though they rank lower overall. 
test                       result  vs the others in its run    difficulty            counts  why
FrontierMath tier 4 v2      92.7%  ▁▂▁▂█▅▂▃●▁▁▁ beats 64/67    very hard · avg 0.1%     13%  sharp
FrontierMath tiers 1–3 v2   90.9%  ▂··▁▁▃▄▄█▆▄● beats 83/86    very hard · avg 1.2%    8.1%  sharp · maxed out
OTIS mock AIME 2024–2025   100.0%  ▃▃▄▃▃▃▇▇█▃▂● beats 187/197  very hard · avg 7.2%    0.1%  sharp · maxed out · noisy
other tests (5) 
test                          result  vs the others in its run    difficulty            counts  why
EBR-bench (Epoch)              78.1%  ▄▂·▆█▂█▆▄▆·● beats 25/25    very hard · avg 0.3%     16%  sharp · at its level
Mystery Game Puzzles (Epoch)   74.0%  █··▂▃▇▄█▂▄●▁ beats 75/77    very hard · avg 9.4%     12%  sharp · at its level · noisy
Furniture Assembly (Epoch)     79.2%  █▆▆▆▆▂▆▂▂▂●▅ beats 29/32    very hard · avg 1.2%     10%  sharp · at its level
SimpleQA Verified              57.3%  ▁▁▃▂▃▃▆▆█●▂▅ beats 64/81    very hard · avg 8.0%    2.1%  at its level
GPQA Diamond                   94.3%  ▁▁▁▁▂▄▆▄▅▇█● beats 203/211  hard · avg 39%          1.3%  sharp · maxed out
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
Epoch AI                                        8   63% ▇▇▇▇▇▇▇▇
LMArena leaderboard dataset                     2   28% ▇▇▇
OpenAI vendor-reported benchmark scores…        1    9% ▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 83.5
Licences: Epoch AI ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) ## How to run it ### Through an API api (OpenAI) · $2.00/M blended · $1 in / $5 out per 1M tokens · [openai.com](https://openai.com/index/introducing-gpt-6-1-terra/) [price source](https://openai.com/index/introducing-gpt-6-1-terra): Standard USD per million uncached input/output tokens; cached input $0.05 excluded. Observed 2026-10-29. [interactive version](https://whatmodel.sh/model/gpt-6-1-terra) · [JSON](https://whatmodel.sh/api?format=json&all=1)