LLM leaderboard
Top 25 of 337 · AA Intelligence Index, then rank on each job's main benchmark
| Model | Intelligence | Chat | Coding | Agents | Writing | Vision | Per task |
|---|---|---|---|---|---|---|---|
57.6max | #161.4 | #259.6 | #569.5 | #11,519 | #187.7 | $5.98 | |
56max | #555 | #163.6 | #371.8 | #151,465 | — | $5.46 | |
55.8 | #853.6 | #557.1 | #471.3 | — | — | — | |
53.4max | #259.1 | #755.1 | #2159.4 | #71,485 | — | $7.63 | |
52.7max | #754.7 | #359.6 | #668.5 | #261,450 | #286.9 | $3.26 | |
52.6high | #357.1 | #457.1 | #177.5 | #21,516 | — | $1.99 | |
52.4max | #1050.1 | #852 | #768.1 | #81,485 | — | $4.52 | |
51.8max | #952.9 | #656.1 | #1066.6 | #191,458 | #386 | $0.72 | |
50.8max | #654.9 | #949 | #2856.6 | #101,471 | #584.7 | $5.86 | |
49.6max | #455.5 | #1142.4 | #3154.1 | #31,501 | — | $8.75 | |
48.1max | #1348.7 | #1933.3 | #2557.9 | #171,461 | #1182 | $1.60 | |
47.6max | #1647.9 | #1043.9 | #1561.7 | #381,438 | #882.9 | $1.04 | |
47max | #1149.5 | #1339.9 | #1860.1 | #141,469 | #683.4 | $1.99 | |
46.4xhigh | #3043.1 | #2525.8 | #1165.6 | #491,424 | #3378.2 | $3.74 | |
46.3thinking | #1249.4 | #1634.8 | #2258.6 | #321,445 | — | $0.13 | |
45.4thinking | #3143.1 | #1438.9 | #2956.2 | — | #982.8 | $5.41 | |
45.1max | #2943.3 | #1733.8 | #966.6 | #581,408 | #1580.6 | $0.26 | |
44.8max | #3742.3 | #1241.9 | #1362.2 | #231,453 | — | $2.01 | |
44.3high | #2544.1 | #2921.2 | #867 | #361,442 | — | $1.48 | |
43.7thinking | #2046.5 | #2033.3 | #3451 | #661,403 | #4076.4 | $1.03 | |
43.6max | #1946.9 | #3412.6 | #2358.3 | #181,461 | #1680.5 | $2.00 | |
43.4max | #2444.4 | #2132.8 | #4736 | — | — | $0.21 | |
42.7thinking | #1548 | #2425.8 | #272.5 | — | #2579.1 | $0.22 | |
42.1max | #3242.9 | #1535.4 | #2059.6 | #471,429 | #1480.7 | $1.40 | |
41.8max | #4639.9 | #2232.8 | #1760.4 | #431,434 | — | $0.28 |
Chat leaderboardCoding leaderboardAgents leaderboardWriting leaderboardVision leaderboardBenchmarks explained
Questions
Which LLM is the best right now?›
Claude Opus 5.5 leads the Artificial Analysis Intelligence Index with 57.6, ahead of Claude Sonnet 5.5 (56.0) and Muse Wonder Full (55.8).
Which LLM is best for chat?›
Claude Opus 5.5 ranks first on Humanity's Last Exam (61.4).
Which LLM is best for coding?›
Claude Sonnet 5.5 ranks first on Terminal-Bench 4.0 (63.6).
Which LLM is best for agents?›
Gemini 4 Argon ranks first on AutomationBench-AA (77.5).
Which LLM is best for writing?›
Claude Opus 5.5 ranks first on Arena Text · Creative writing (1,519).
Which LLM is best for vision?›
Claude Opus 5.5 ranks first on MMMU-Pro (87.7).
What is the best open-weights LLM?›
MiMo v2.6 Pro (46.3), then DeepSeek V4.1 Flash 1017 (45.1), then GLM 5.3 (44.8) on the Intelligence Index.