Anthropic: Claude Opus 4.5

What the data says

Benchmarked on 22 of 27 tracked axes. Every figure below is the model's own published result on that board.

Uneven: 64.8 percentile points between its best and worst axis — 100 pct on BFCL v4 (function calling) versus 35.2 pct on HHEM Hallucination (Vectara).

• BFCL v4 (function calling)Dormant board100 pct77.5%Verify →
• SWE-bench Verified (live board)100 pct79.2%Verify →
• MMLU-Pro (Artificial Analysis)99.6 pct89.5%Verify →
• LiveCodeBench (Artificial Analysis)97.7 pct87.1%Verify →
• Instruction following94.4 pct1,474Verify →
• Coding93.4 pct1,523Verify →
• Hard prompts92.2 pct1,498Verify →
• Overall89.6 pct1,470Verify →
• SimpleBench78.2 pct62.0%Verify →
• Terminal-Bench77.3 pct63.1%Verify →
• Factuality Arena75.6 pct1,459Verify →
• SWE-bench Verified75 pct76.7%Verify →
• GPQA Diamond70.6 pct86.0%Verify →
• IFBench (Artificial Analysis)67.6 pct58.0%Verify →
• Humanity's Last Exam67.4 pct25.2%Verify →
• AIME (math)63.5 pct86.1%Verify →
• WebDev Arena61.2 pct1,494Verify →
• FrontierMath58 pct20.7%Verify →
• Document Arena57.9 pct1,465Verify →
• ARC-AGI-257.3 pct37.6%Verify →
• SimpleQA Verified53.7 pct45.7%Verify →
• HHEM Hallucination (Vectara)35.2 pct89.1%Verify →
• Vision ArenaNot benchmarked
• Search ArenaNot benchmarked
• Agent ArenaNot benchmarked
• LiveBenchNot benchmarked
• Terminal-Bench 4.0Not 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.

How to run it

curl https://openrouter.ai/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"anthropic/claude-opus-4.5","messages":[{"role":"user","content":"Hello"}]}'

Identifiers, limits and prices are synced from the upstream source shown above. Any field we haven't synced is left out entirely rather than guessed.