Atlas

Compare Models

Claude Opus 5.5 vs GPT-6.1 Terra

Claude Opus 5.5 leads by 2.4 capability points (≈95% interval −0.2 to +5.0): statistically tied · P(higher) ≈ 97%

Wins 67 of 87 shared benchmarks (19 lost, 1 tied)

Capability vs. relative cost
Claude Opus 5.5GPT-6.1 TerraClosed weightsOpen weights

Each line runs through one model’s priced reasoning settings, cheapest to most expensive; grey points are other models’ priced settings in the same range. Results whose reasoning effort the source didn’t report are left out wherever a model has named effort levels. Capability and relative cost are the published estimates (relative cost is cost per task as a multiple of GPT-5 (high)); hover a setting, or use the arrow keys, to set it beside the other model’s same setting.

Claude Opus 5.5vsGPT-6.1 Terracapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins67 – 19 · 87 shared, 1 tied

Comparing matching setups only (24 benchmarks excluded) · suites counted once
← Claude Opus 5.5GPT-6.1 Terra →

Claude Opus 5.5vsGPT-6.1 Terrawins by categoryData Nov 15, 2026Data: benchmarks.bio, Artificial Analysis, Epoch AI and 29 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-6.1 Terra minus Claude Opus 5.5.

BioSecBench-Function V1 Clean RerunsScience73.639.0−34.6
TxBench Antibody Discovery V1 (100 evaluations)Science8.340.7+32.3
BioSecBench-Refusal V1 Clean RerunsSafety3.834.2+30.4
TxBench Preclinical Pharmacology V1 Clean RerunsScience26.355.7+29.3
TxBench Oligo Discovery V1 (113 evaluations)Science1.530.4+28.9
FrontierMemory v1.1Agents73.346.7−26.7
AA-Briefcase v1.1 EloAgents18011401−21.9
VoxelBench Text RatingGames26132313−21.4
SpatialBench-Long V1 Clean RerunsScience20.240.5+20.2
Terminal-Bench-Science 0.1Science61.941.9−20.0
Toloka Arena (v3)Agents75.055.0−20.0
SWE Atlas v1.1 - OverallCode77.457.7−19.7
TaxCalcBench TY25Finance & Accounting42.024.0−18.0
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.037.1−17.9
GDPval-AA v2.1Professional Work18661476−17.2
SpireClimbGames5.822.3+16.5
BioSecBench-Surveillance V1 Clean RerunsScience55.839.9−15.9
VariantBench V1 Clean RerunsScience50.634.8−15.8
EpiBench V1 Clean RerunsScience39.023.3−15.7
HardGeoBench <=100miMultimodal61.346.3−15.0
SimpleQA Verified (Epoch anti-abstention)General QA72.257.3−14.9
Terminal-Bench 4.0Code64.850.3−14.5
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science25.411.2−14.2
LMArena Agent Arena Net Improvement (Steering Burden)Agents19.512.0−14.2
Design Arena Overall Code CategoriesCode13841300−13.9
FrontierCode 1.1 Main (Artificial Analysis)Code57.143.7−13.3
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine29.316.0−13.3
Terminal-Bench 4.0 (Artificial Analysis)Code59.646.5−13.1
HardGeoBench Avg DistanceMultimodal6691091−13.0
ScBench-Long V1 Clean RerunsScience6.818.8+11.9
SciCode 1.0.1 (Artificial Analysis)Code66.955.1−11.8
Humanity's Last Exam (Text-Only)General QA57.546.2−11.3
Visa Agentic Commerce BenchmarkAgents62.251.1−11.1
LMArena Agent Arena Praise vs ComplaintAgents40.320.0−10.1
DistressBenchSafety75.665.9−9.7
MacroscopeBenchCode73.182.3+9.2
Blueprint-Bench 2Multimodal0.50.4−9.1
HiL-Bench (Human-in-Loop Benchmark)Agents59.068.0+9.0
GameBench 2Games94.885.8−8.9
HardGeoBench Country AccuracyMultimodal67.558.8−8.7
AA-Omniscience IndexIndexes46.429.8−8.3
LLM Chess EloGames15751391−7.6
HardGeoBench <=10miMultimodal42.535.0−7.5
HLAB-Vals-1.1Legal9.21.7−7.5
Humanity's Last Exam DiamondGeneral QA54.647.3−7.3
LiveBench 2.0Agents86.579.7−6.9
EBR-bench v4 (card ban, single agent)Games71.478.1+6.7
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.84.5−6.3
ChessBench EloGames13491251−5.8
LMArena Agent Arena Confirmed SuccessAgents20.79.7−5.5
FORTRESS Average Risk Score (ARS)Safety13.218.2−5.0
ChessBench CoherenceGames80.184.8+4.7
ALE-BenchCode27482875+4.5
BuseyBench SVGMultimodal6.87.3+4.3
Furniture Assembly BenchmarkMultimodal83.379.2−4.2
FrontierCode 2.0 (Artificial Analysis)Code46.442.3−4.1
DualEntry Accounting AI BenchmarkFinance & Accounting82.278.2−4.0
CritPt (avg@5)Science31.728.0−3.7
GBENCH One-shot Coding GScoreGames73.977.6+3.7
GDPxProfessional Work12.08.5−3.5
BioSecBench-Structure V1 (58 evaluations)Science29.325.9−3.4
Agents' Last Exam Overall ScoreAgents63.260.0−3.2
Mystery Game PuzzlesGames71.074.0+3.0
ScBench V1 Clean RerunsScience55.458.3+2.9
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine27.830.6+2.8
Excel Modeling Benchmark v1.1Finance & Accounting75.873.3−2.5
FrontierMath Tier 4 (v2) - PrivateMath95.092.7−2.3
DAXBench v2Code95.793.6−2.1
AA-LCR v1.1Search84.782.7−2.0
APEX-Agents 1.1Professional Work81.379.3−2.0
MMMU ProMultimodal86.684.7−1.9
APEX-Accounting 1.1Finance & Accounting65.563.7−1.8
APEX-1 (Mercor, mean score)Professional Work65.864.2−1.6
ChessBenchGames65.364.0−1.3
ChessBench AccuracyGames79.378.3−1.1
LMArena Agent Arena Steering BurdenAgents18.616.6−1.0
LMArena Agent Arena Bash ReliabilityAgents15.113.4−0.9
MetagenomicsBench V1Science39.338.7−0.7
SpatialBench V1 Clean RerunsScience72.873.3+0.6
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting86.285.7−0.5
Epoch Capabilities Index (ECI)Indexes169168−0.5
FrontierMath Tiers 1-3 (v2) - PrivateMath91.290.9−0.4
CritPt v2 (agentic, avg@5)Science81.781.4−0.3
Prophet Arena Arena Score (Fixed Context)Forecasting85.585.2−0.3
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.828.6−0.2
LMArena Agent Arena Tool HallucinationAgents0.71.0+0.2
OTIS Mock AIME 2024-2025Math100.0100.00.0

Claude Opus 5.5vsGPT-6.1 Terranormalised score marginsData Nov 15, 2026Data: benchmarks.bio, Artificial Analysis, Epoch AI and 29 othersVillage LLM Atlasatlas.village.dev

Win counts use each benchmark’s declared score direction. Represented suite slices and the derived Capability score are excluded from headline totals. Missing or incompatible setups are excluded, never counted as losses.
Loading Atlas data…