GPT 4 5 Preview

What the data says

Benchmarked on 11 of 27 tracked axes. Every figure below is the model's own published result on that board.

Uneven: 65.6 percentile points between its best and worst axis — 81.3 pct on Instruction following versus 15.7 pct on ARC-AGI-2.

• Instruction following81.3 pct1,437Verify →
• Overall80.6 pct1,445Verify →
• LiveBench79.4 pct68.9%Verify →
• Hard prompts69.7 pct1,440Verify →
• Coding66.8 pct1,458Verify →
• Vision Arena56 pct1,225Verify →
• GPQA Diamond50.2 pct68.7%Verify →
• AIME (math)35 pct37.8%Verify →
• SimpleBench27.6 pct34.5%Verify →
• Humanity's Last Exam17.4 pct5.4%Verify →
• ARC-AGI-215.7 pct0.8%Verify →
• WebDev ArenaNot benchmarked
• Factuality ArenaNot benchmarked
• Search ArenaNot benchmarked
• Document ArenaNot benchmarked
• Agent ArenaNot benchmarked
• FrontierMathNot benchmarked
• SimpleQA VerifiedNot benchmarked
• BFCL v4 (function calling)Not benchmarked
• SWE-bench Verified (live board)Not benchmarked
• Terminal-Bench 4.0Not benchmarked
• MMLU-Pro (Artificial Analysis)Not benchmarked
• LiveCodeBench (Artificial Analysis)Not benchmarked
• IFBench (Artificial Analysis)Not benchmarked
• HHEM Hallucination (Vectara)Not benchmarked
• SWE-bench VerifiedNot benchmarked
• Terminal-BenchNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.