OpenAIUnited States· released Jul 2026

GPT-5.6 Terra

Name: GPT-5.6 Terra
Author: OpenAI

Ranked #6, 18.1% behind the frontier on average.

OpenAI, United States. One of 26 models from this lab on the Index.

Third-party publishedMedium confidence (7+ comparable tests, 3+ Index categories)

Available in Theo as Theo Balanced

Theo orchestrates it for the steps it does best and always shows which engine answered.

Use it in Theo

Snapshot September 8, 2026

OpenCharts Index

81.9

Ranked #6 of 36 · 18.1% behind the frontier on average · 18 comparable tests · 3 counting categories

Drop any single test and the Index lands between 80.2 and 85.3.

Needs more results (3 of 8 so far)

Averaging 84.2 on the 3 it has; the score opens at 4 of 8.

Distance to today's frontier on the hardest tests. Not a measure of intelligence.

What it's known for

GPT-5.6 Terra, in 18 results.

Strongest in reasoning (87.1) and math (84.5). It trails the frontier most in coding, 25.9% behind on average. No comparable result yet in long context. Between 15% and 20% behind the best published results, on average.

Strongest category

Reasoning

Hard, novel problems: graduate science, abstract puzzles, expert exams.

87.1

ARC-AGI-196.5% · 98% of best
GPQA Diamond93.3% · 97% of best
CritPt30.0% · 93% of best
ARC-AGI-283.9% · 88% of best
SimpleBench48.9% · 60% of best

Mean of 5 of the category's 6 comparable tests. Counts toward the Index.

Best single result

Mock AIME 2024–2025

Solve competition mathematics problems that have a single exact integer answer.

99.7%

99.7% of the best (100.0%) · #5 of 78.

Olympiad-style problems need long, careful chains of reasoning, and there is no partial credit.

Where it trails

Code Arena (WebDev)

Build a web app from the same prompt as a rival model; real users vote blind on the result.

1,520

66.3 points short of parity with the board leader (1,797) · #28 of 86.

People judging finished apps side by side is the most honest measure of front-end quality.

Not measured yet

No comparable result yet in this category. A blank is a blank, never a zero, and it does not lower the Index.

Long context

Covered, but not averaged into the Index:

Knowledgesingle-test category, shown beside the Index · 57.1
Agentic1 of 8 tests · needs 4 · mean 86.5
Multimodalsingle-test category, shown beside the Index · 86.5
Human preferencesingle-test category, shown beside the Index · 88.3

Every score, with its source

See it at work

What GPT-5.6 Terra was asked to do, and how it did.

Every test it has taken, by category, with the kind of task it faced, who produced the number and where the result landed against the best published one. The examples are original and illustrative, never items from the datasets themselves.

Human preference · mean 88.3 across 1 comparable test · a single-test category, shown beside the Index and never averaged into it

Text ArenaLeaderboard

Answer the same prompt as a rival model; real users vote blind on which answer they prefer.

1,466

44% expected win rate against the board leader (1,507)
#33 of 92

xhigh · board 2026-09-02

Source: Arena

Every result is read against the best published one on its test: percent scores as a share of the best above chance, Arena ratings as the expected win rate against the board leader, open-ended values on a log scale. Sitting a harder exam never lowers a model. Every number links to the publisher that produced it.

Head to head

GPT-5.6 Terra against whoever you pick.

Choose a rival. Every counted test both have taken appears side by side as a share of the best published result on that test, with the category means above. A win is a higher share; a gap under 2 points is a tie, because one item on a 45-problem exam is 2.2 points and publishers' own standard errors are of that order.

GPT-5.6 Terra

#6·Index 81.9B+

0 : 12

wins · 1 tie · 13 shared

GPT-6 Astra

#1·Index 99.0A+

GPT-6 Astra comes out ahead on 12 of the 13 tests both have taken.

Category means

Reasoning
87.1
99.6
Knowledge
57.1
100.0
Coding
74.1
97.7
Math
84.5
99.8
Agentic
86.5
98.5
Multimodal
86.5
—
Human preference
88.3
—

Shared tests · biggest gaps first

Code Arena (WebDev)Coding
1,520
1,797
−66.3
SimpleQA VerifiedKnowledge
43.2%
75.6%
−42.9
FrontierMath Tier 4Math
70.7%
97.6%
−27.5
ProofBenchMath
74.0%
99.0%
−25.0
FrontierCodeCoding
41.3%
53.3%
−22.4
WeirdMLCoding
78.3%
92.9%
−15.7
ARC-AGI-2Reasoning
83.9%
95.0%
−11.7
FrontierMath (Tiers 1–3)Math
86.0%
93.7%
−8.2
CritPtReasoning
30.0%
31.7%
−5.3
SciCodeCoding
53.9%
56.5%
−4.1
GPQA DiamondReasoning
93.3%
95.8%
−3.5
ARC-AGI-1Reasoning
96.5%
98.5%
−2.0
Mock AIME 2024–2025Math
99.7%
100.0%
tie

Bars are distances to the best published result on each test (percent scores above chance, Arena ratings as win rate against the leader, open-ended values on a log scale). Raw values as the publishers report them.

Open GPT-6 Astra

Scores

Every test, every source

The best published run per benchmark, the raw value as the publisher reports it, who produced the number, how close it comes to the best published result on that test, and where the model sits among every model scored on it. Percent scores are read above chance, Arena ratings as a win rate against the board leader, open-ended values on a log scale. A test with too few models to compare against is shown but not counted.

Category means

Reasoning87.1 · 5/6
Knowledgebeside57.1 · 1/1
Coding74.1 · 5/7
Math84.5 · 4/5
Agentic86.5 · 1/8
Multimodalbeside86.5 · 1/1
Human preferencebeside88.3 · 1/1
Long context—

Mean · tests taken of the category's comparable tests. A faded category either rests on fewer than half of them or is a single-test category (marked “beside”), which is shown next to the Index and never averaged into it.

GPQA DiamondEpoch-runReasoning

93.3%

96.5% of the best (95.8%), above chance
87th percentile of 78 · n = 198

max

GPT-5.6 Terra

GPT-5.6 Terra, in 18 results.

What GPT-5.6 Terra was asked to do, and how it did.

GPT-5.6 Terra against whoever you pick.

Every test, every source

Compare with its neighbours.

The best models, ranked here, working inside Theo.