Atlas

Compare Models

Opus 4.8 vs GPT-6.1 Terra

GPT-6.1 Terra leads by 5.6 capability points (≈95% interval +3.3 to +7.8): a clear lead over Opus 4.8 · P(higher) ≈ 100%

Wins 51 of 63 shared benchmarks (11 lost, 1 tied)

Capability vs. relative cost
Opus 4.8GPT-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.

Opus 4.8vsGPT-6.1 Terracapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins11 – 51 · 63 shared, 1 tied

Comparing matching setups only (7 benchmarks excluded) · suites counted once
← Opus 4.8GPT-6.1 Terra →

Opus 4.8vsGPT-6.1 Terrawins by categoryData Nov 15, 2026Data: benchmarks.bio, Artificial Analysis, Epoch AI and 18 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 Opus 4.8.

ExploitBench - release-reported exploit progressCybersecurity40.097.1+57.1
VoxelBench Text RatingGames16272313+48.9
LLM Chess EloGames4501391+38.6
Mystery Game PuzzlesGames36.074.0+38.0
Furniture Assembly BenchmarkMultimodal42.579.2+36.7
FrontierMath Tier 4 (v2) - PrivateMath56.192.7+36.6
Terminal-Bench 4.0 (Artificial Analysis)Code21.755.6+33.8
HiL-Bench (Human-in-Loop Benchmark)Agents35.368.0+32.7
Blueprint-Bench 2Multimodal0.10.4+27.6
ALE-BenchCode21022875+27.4
Terminal-Bench 4.0Code23.249.5+26.3
Humanity's Last Exam DiamondGeneral QA23.147.3+24.2
GBENCH One-shot Coding GScoreGames54.077.6+23.5
GameBench 2Games62.585.8+23.4
BuseyBench SVGMultimodal5.27.3+20.7
Legal Research BenchLegal43.826.0−17.8
ChessBench EloGames9611251+17.3
NoCoT-Bench NCRI 15.2General QA102130+15.6
APEX-Agents 1.1Professional Work64.479.3+14.9
Agents' Last Exam Overall ScoreAgents45.260.0+14.8
AutomationBench-AA 1.0.6Agents45.660.2+14.6
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine19.430.6+11.1
FrontierMath Tiers 1-3 (v2) - PrivateMath80.090.9+10.9
HardGeoBench <=100miMultimodal36.346.3+10.0
Tax Agent BenchFinance & Accounting64.955.0−9.9
HealthBench Professional (Length-Adjusted)Health & Medicine53.062.7+9.7
MetagenomicsBench V1Science48.338.7−9.7
Artificial Analysis Progress Index v1Indexes318345+9.4
Design Arena Overall Code CategoriesCode12521300+8.0
VariantBench V1 Clean RerunsScience42.434.8−7.6
SpatialBench V1 Clean RerunsScience65.873.3+7.5
HardGeoBench <=10miMultimodal27.535.0+7.5
TxBench Antibody Discovery V1 (100 evaluations)Science33.340.7+7.3
SpatialBench-Long V1 Clean RerunsScience33.340.5+7.1
TxBench Oligo Discovery V1 (113 evaluations)Science37.530.4−7.1
BioSecBench-Surveillance V1 Clean RerunsScience46.939.9−7.0
Epoch Capabilities Index (ECI)Indexes160168+6.6
ChessBench AccuracyGames72.478.3+6.3
Artificial Analysis Intelligence Index v4.3.2Indexes41.848.1+6.3
GPQA DiamondGeneral QA88.494.3+5.9
GDP.pdf (Artificial Analysis, All-pass)Professional Work22.828.6+5.8
DistressBenchSafety71.465.9−5.5
NoCoT-Bench NCKI (kspine v3)General QA117110−5.2
ChessBench CoherenceGames79.784.8+5.1
HardGeoBench Country AccuracyMultimodal53.858.8+5.0
HardGeoBench Avg DistanceMultimodal12451091+4.7
AA-Briefcase v1.1 EloAgents13201401+4.4
SimpleQA Verified (Epoch anti-abstention)General QA53.057.3+4.3
BioSecBench-Function V1 Clean RerunsScience43.139.0−4.0
TaxCalcBench TY25Finance & Accounting20.024.0+4.0
LiveBench 2.0Agents76.879.7+2.9
EpiBench V1 Clean RerunsScience25.823.3−2.5
ScBench-Long V1 Clean RerunsScience16.518.8+2.3
TxBench Preclinical Pharmacology V1 Clean RerunsScience57.755.7−2.0
OTIS Mock AIME 2024-2025Math98.3100.0+1.7
APEX-1 (Mercor, mean score)Professional Work62.764.2+1.5
Humanity's Last Exam (Text-Only)General QA44.846.2+1.4
Vibe Code Bench v1.1Code82.784.0+1.3
ScBench V1 Clean RerunsScience57.158.3+1.2
SciCode 1.0.1 (Artificial Analysis)Code54.455.1+0.7
GDPval-AA v2.1Professional Work14611476+0.6
AA-Omniscience IndexIndexes28.829.8+0.5
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine16.016.00.0

Opus 4.8vsGPT-6.1 Terranormalised score marginsData Nov 15, 2026Data: benchmarks.bio, Artificial Analysis, Epoch AI and 18 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…