GPT-6.1 Terra (max): benchmark scores, price and ranking

GPT-6.1 Terra (max) from OpenAI ranks #10 of 827 LLMs on the LLMs Tiger aggregate benchmark score (59.8), based on 13 benchmark results captured up to 2026-11-18.

Organisation
OpenAI
Type
Frontier model
Open weights
No
Released
2026-10-26 (source: epoch)
Cheapest API price
$1.00 in / $5.00 out per million tokens (via azure, observed 2026-11-18T15:47Z)
Aggregate rank
#10 of 827 (score 59.8)

Relative to other models, GPT-6.1 Terra (max) is strongest on ALE-Bench (#3 of 93) and weakest on SciCode (#26 of 113).

Across the 13 benchmarks where at least ten models are scored, GPT-6.1 Terra (max) is in the top 10% on 5 and in the top half on 13.

Its cheapest listed input price, $1.00 per million tokens, is lower than that of 35% of the 443 models with a listed price.

GPT-6.1 Terra (max) benchmark results

GPT-6.1 Terra (max) benchmark scores, ranked against all models measured on each benchmark
BenchmarkScoreRankSourceCaptured
ALE-Bench2,609#3 of 93Epoch AI Benchmarking Hub2026-11-18
GPQA Diamond94.3%#11 of 302Epoch AI Benchmarking Hub2026-11-18
OTIS Mock AIME100.0%#10 of 206Epoch AI Benchmarking Hub2026-11-18
Mystery Games74.0%#5 of 69Epoch AI Benchmarking Hub2026-11-18
FrontierMath T1-390.9%#6 of 78Epoch AI Benchmarking Hub2026-11-18
CritPt28.0%#14 of 138Epoch AI Benchmarking Hub2026-11-18
FrontierMath T492.7%#7 of 60Epoch AI Benchmarking Hub2026-11-18
EBR-Bench78.1%#4 of 27Epoch AI Benchmarking Hub2026-11-18
Furniture Assembly79.2%#6 of 36Epoch AI Benchmarking Hub2026-11-18
MirrorCode76.7%#6 of 35Epoch AI Benchmarking Hub2026-11-18
APEX-Agents68.3%#4 of 23Epoch AI Benchmarking Hub2026-11-18
SimpleQA Verified57.3%#17 of 83Epoch AI Benchmarking Hub2026-11-18
SciCode55.1%#26 of 113Epoch AI Benchmarking Hub2026-11-18

About GPT-6.1 Terra (max)

GPT-6.1 Terra (max) is a frontier model from OpenAI. It was released on 2026-10-26, according to Epoch AI. The cheapest API listing we track charges $1.00 per million input tokens and $5.00 per million output tokens, through azure. We hold 13 benchmark results for it across 4 areas, and 5 are in the top tenth of their boards.

Where GPT-6.1 Terra (max) stands

On our aggregate score, GPT-6.1 Terra (max) is #10 of 827 ranked models, in the top ten. Its score is 59.8; the scale is centred on 50, so it sits 9.8 points above the typical model. That beats 99% of the ranked field. Among the 91 ranked OpenAI models it is #4. The next one up is GPT-6.1 Sol (max) at 60.4. The next one down is GPT-6 Sol (max) at 58.0.

GPT-6.1 Terra (max), benchmark by benchmark

On ALE-Bench, GPT-6.1 Terra (max) scores 2,609 and ranks #3 of 93. Only 2 models score higher, led by GPT-6 Astra (max) at 2,951, which is 342 points ahead. It covers AtCoder heuristic-optimisation contests (performance rating). The board median is 800, so GPT-6.1 Terra (max) is 1,809 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

On GPQA Diamond, GPT-6.1 Terra (max) scores 94.3% and ranks #11 of 302. Only 10 models score higher, led by GPT-6.2 Astra (max) at 96.8%, which is 2.5 points ahead. It covers expert-level science questions. The board median is 75.1%, so GPT-6.1 Terra (max) is 19.2 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

On OTIS Mock AIME, GPT-6.1 Terra (max) scores 100.0% and ranks #10 of 206. Only 9 models score higher, led by Claude Fable 5.1 (max) at 100.0%, which is 0.0 points ahead. It covers competition maths (AIME 2024-25 style). The board median is 66.4%, so GPT-6.1 Terra (max) is 33.6 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

On Mystery Games, GPT-6.1 Terra (max) scores 74.0% and ranks #5 of 69. Only 4 models score higher, led by GPT-6.2 Astra (max) at 95.0%, which is 21.0 points ahead. It covers deductive reasoning puzzles. The board median is 26.0%, so GPT-6.1 Terra (max) is 48.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

On FrontierMath T1-3, GPT-6.1 Terra (max) scores 90.9% and ranks #6 of 78. Only 5 models score higher, led by Claude Fable 5.2 (max) at 95.1%, which is 4.2 points ahead. It covers research-level maths, tiers 1-3. The board median is 61.6%, so GPT-6.1 Terra (max) is 29.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

GPT-6.1 Terra (max) manages 28.0% on CritPt, good for #14 of 138. Gemini 4 Carbon (high) leads at 36.0%, 8.0 points further up, and 90% of models sit below. The results either side are Muse Wonder (max) (28.3%) and gemini-4-argon-high (27.1%). It covers unpublished research-level physics problems. The board median is 2.0%, so GPT-6.1 Terra (max) is 26.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

MirrorCode puts GPT-6.1 Terra (max) at 76.7%, #6 of 35. That beats 83% of the models tested, but 21.1 points short of Claude Sonnet 5.5 (max) (97.8%). The results either side are Claude Haiku 5.5 (max) (78.9%) and Claude Fable 5 (max) (70.0%). It covers reimplementing programs from behaviour. The board median is 48.9%, so GPT-6.1 Terra (max) is 27.8 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

On APEX-Agents, GPT-6.1 Terra (max) reaches 68.3%. It ranks #4 of 23, ahead of 83% of the field, and 7.2 points behind Claude Sonnet 5.5 (max) at 75.5%. The results either side are DeepSeek V4.1 Flash 1017 (max) (69.8%) and Muse Wonder (max) (66.1%). It covers long professional agent tasks. The board median is 51.6%, so GPT-6.1 Terra (max) is 16.7 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

GPT-6.1 Terra (max) manages 57.3% on SimpleQA Verified, good for #17 of 83. Claude Fable 5.2 (max) leads at 94.0%, 36.7 points further up, and 80% of models sit below. The results either side are Claude Opus 5 (max) (59.9%) and Muse Spark Open (xhigh) (56.6%). It covers short factual questions. The board median is 41.4%, so GPT-6.1 Terra (max) is 15.9 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

SciCode puts GPT-6.1 Terra (max) at 55.1%, #26 of 113. That beats 77% of the models tested, but 12.6 points short of Claude Fable 5.2 (max) (67.7%). The results either side are GPT-5.5 (xhigh) (55.8%) and Claude Haiku 5.5 (max) (55.0%). It covers scientific research code. The board median is 46.1%, so GPT-6.1 Terra (max) is 9.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-11-18.

The full table above also lists the other 3 results, including boards with fewer than ten models, where a rank says little.

Strengths and weaknesses

Software engineering: strong, ahead of 87% of the field on average across 2 benchmarks. Its best board here is ALE-Bench (#3 of 93) and its weakest is SciCode (#26 of 113). Reasoning, maths and science: strong, ahead of 92% of the field on average across 5 benchmarks. Its best board here is GPQA Diamond (#11 of 302) and its weakest is FrontierMath T4 (#7 of 60). General knowledge and chat: strong, ahead of 85% of the field on average across 5 benchmarks. Its best board here is Mystery Games (#5 of 69) and its weakest is SimpleQA Verified (#17 of 83).

Price and value

The median input price across the 443 priced models is $0.482 per million tokens, so GPT-6.1 Terra (max) costs more than the typical listing. We found no cheaper model with a similar or better aggregate score, so GPT-6.1 Terra (max) is the lowest-priced option at its level. Similar scores cost more on Claude Sonnet 5.5 (max) ($2.00) and GPT-6.1 Sol (max) ($2.00).

Release history

The OpenAI model released before it was GPT-6.1 Sol (max) on 2026-09-29, which scores 60.4. The next one was GPT-6.2 Astra (max) on 2026-11-17, at 60.5.

Reading these numbers

The 13 scores on this page come from 1 source: Epoch AI Benchmarking Hub (13). No official leaderboard supplied any of them, so treat them as third-party measurements. The leaderboard also lists 7 other reasoning-effort settings of this model, which redirect here. This page shows the highest.

GPT-6.1 Terra (max) is also listed at 7 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.

Models ranked nearby

More from OpenAI

Compare it on the live leaderboard → · How the aggregate score works