GPT-6.2 Astra vs Qwen4-Max
GPT-6.2 Astra leads by 4.9 capability points (≈95% interval +2.2 to +7.6): a clear lead over Qwen4-Max · P(higher) ≈ 100%
Wins 44 of 54 shared benchmarks (10 lost)
Capability vs. relative cost
GPT-6.2 AstraQwen4-MaxClosed weightsOpen weights
Benchmark wins44 – 10 · 54 shared
Comparing matching setups only (3 benchmarks excluded) · suites counted once
Area← GPT-6.2 AstraQwen4-Max →
Data: Artificial Analysis, Surge AI, Vals AI and 7 othersatlas.village.dev
Margins
Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Qwen4-Max minus GPT-6.2 Astra.
| KradleBench | Games | 78.3 | 11.0 | −67.3 |
| Terminal-Bench-Science 0.1 | Science › Multidisciplinary | 78.6 | 23.3 | −55.2 |
| Humanity's Last Exam Diamond | Knowledge & reasoning › General knowledge | 72.2 | 44.9 | −27.3 |
| Code Migration v1.1 (Vals Index v2.2 subset) | Software › Software engineering | 83.6 | 57.3 | −26.3 |
| EnterpriseBench: CoreCraft | Tool & computer use › Customer service & operations | 43.6 | 68.7 | +25.1 |
| DAYJOB: Law (November 2026 release) | Professional › Legal | 41.4 | 17.5 | −23.9 |
| EnterpriseOps-Gym-AA | Tool & computer use › Customer service & operations | 25.3 | 48.5 | +23.2 |
| sudo L7 (October 2026 rescoring) | Software › Software engineering | 38.3 | 15.6 | −22.7 |
| FrontierCode 2.0 (Artificial Analysis) | Software › Software engineering | 49.0 | 26.5 | −22.5 |
| DAYJOB: Finance (October 2026 rescoring) | Professional › Finance, accounting & tax | 45.8 | 25.3 | −20.5 |
| Medical Long Context Reasoning (MLCR-AA, October 2026 methodology) | Health & medicine | 39.4 | 19.4 | −20.0 |
| DAYJOB: Accounting | Professional › Finance, accounting & tax | 42.0 | 22.4 | −19.6 |
| Artificial Analysis Intelligence Index v5.0.1 | Composite indexes | 62.1 | 42.9 | −19.2 |
| CritPt v2 (agentic, avg@5) | Science › Physical sciences | 83.1 | 66.0 | −17.1 |
| Artificial Analysis Engineering Index | Science › Engineering | 66.6 | 50.1 | −16.6 |
| AA-AnalystAgent (pass^5) | Professional › General knowledge work | 62.5 | 46.3 | −16.2 |
| Terminal-Bench 4.0 (Artificial Analysis) | Software › Software engineering | 64.1 | 48.5 | −15.7 |
| LMArena Agent Arena Net Improvement (Steering Burden) | Tool & computer use › General agents | 15.2 | 7.3 | −15.5 |
| Chartography (November 2026 rescoring) | Perception & media › Documents & charts | 90.7 | 75.3 | −15.4 |
| FrontierCUA v1 pass@1 | Tool & computer use › Computer & browser use | 32.5 | 17.5 | −15.0 |
| Terminal-Bench 4.0 | Software › Software engineering | 63.1 | 48.5 | −14.6 |
| Tax Agent Bench | Professional › Finance, accounting & tax | 57.9 | 70.0 | +12.2 |
| LMArena Agent Arena Praise vs Complaint | Tool & computer use › General agents | 37.0 | 13.0 | −12.0 |
| LiveBench 2.0 (2026-11-06 release) | Composite indexes | 84.9 | 73.1 | −11.8 |
| AA-Omniscience v1.1 Index | Knowledge & reasoning › General knowledge | 50.0 | 27.7 | −11.1 |
| ComplexConstraints (November 2026 rescoring) | Language & writing › Instruction following | 69.6 | 59.1 | −10.5 |
| DAYJOB: Healthcare (October 2026 rescoring) | Health & medicine | 34.4 | 24.4 | −10.0 |
| Artificial Analysis Economics Index | Composite indexes | 61.1 | 51.3 | −9.8 |
| Artificial Analysis Legal Index | Professional › Legal | 58.7 | 49.6 | −9.2 |
| Artificial Analysis Healthcare & Medical Index | Health & medicine | 57.0 | 48.2 | −8.8 |
| Artificial Analysis Finance & Accounting Index | Professional › Finance, accounting & tax | 60.9 | 54.1 | −6.8 |
| Artificial Analysis Strategy & Ops Index | Professional › General knowledge work | 63.3 | 56.9 | −6.4 |
| HANDBOOK.md Agents v1.1 | Tool & computer use › Customer service & operations | 45.8 | 39.4 | −6.4 |
| GDP.xlsx | Professional › General knowledge work | 29.7 | 23.7 | −6.0 |
| Finance Agent v2.1 | Professional › Finance, accounting & tax | 60.7 | 66.7 | +5.9 |
| GDP.pdf (November 2026 rescoring) | Professional › General knowledge work | 49.2 | 43.6 | −5.6 |
| Excel Modeling Benchmark v1.1 | Professional › Finance, accounting & tax | 76.0 | 70.8 | −5.2 |
| AA-Briefcase v1.1 Elo | Professional › General knowledge work | 1594 | 1686 | +5.0 |
| DualEntry Accounting AI Benchmark | Professional › Finance, accounting & tax | 78.2 | 83.2 | +5.0 |
| HLAB-Vals-1.1 | Professional › Legal | 16.3 | 21.3 | +5.0 |
| AutomationBench-AA 1.0.6 | Tool & computer use › Customer service & operations | 77.2 | 72.6 | −4.7 |
| Vals Finance Index | Professional › Finance, accounting & tax | 57.3 | 62.0 | +4.7 |
| Vals Index v2.2 | Composite indexes | 67.7 | 63.2 | −4.5 |
| GDP.pdf (Artificial Analysis, All-pass) | Professional › General knowledge work | 29.8 | 26.2 | −3.6 |
| Vibe Code Bench v1.1 | Software › Frontend & design | 89.9 | 86.7 | −3.2 |
| LMArena Agent Arena Steering Burden | Tool & computer use › General agents | 14.0 | 8.0 | −3.0 |
| LMArena Agent Arena Bash Reliability | Tool & computer use › General agents | 7.7 | 3.4 | −2.2 |
| LMArena Agent Arena Confirmed Success | Tool & computer use › General agents | 12.4 | 8.8 | −1.8 |
| Legal Research Bench | Professional › Legal | 42.8 | 41.3 | −1.4 |
| Vals Legal Index | Professional › Legal | 69.7 | 70.7 | +1.0 |
| Agents' Last Exam v1.1 (ALE v1.1) | Professional › General knowledge work | 47.5 | 46.7 | −0.8 |
| LLM Debate Benchmark | Language & writing › Language & wordplay | 1673 | 1679 | +0.8 |
| DAYJOB: Tax | Professional › Finance, accounting & tax | 9.2 | 8.8 | −0.4 |
| LMArena Agent Arena Tool Hallucination | Tool & computer use › Tool calling | 1.2 | 1.2 | 0.0 |
Data: Artificial Analysis, Surge AI, Vals AI and 7 othersatlas.village.dev