⌕Is ___ SOTA?
benchmarks / lmarena-factuality

LMArena Text factuality

Arena score. Status active. Contamination risk low. Maintainer LMArena. Source.

ModelScoreDateKind
Claude Opus 4.6 Thinking1492.66894426350382026-08-11snapshot
Claude Fable 51484.030845013962026-09-13snapshot
Claude Opus 4.61489.05634626882562026-10-08snapshot
GPT-5.41484.56049908247592026-10-08snapshot
GPT-5.51486.0122265957032026-10-08snapshot
Gemini 3 Pro1481.22876693633542026-10-08snapshot
GPT-5.51482.14490974412272026-10-08snapshot
Qwen3 7 Max Preview1480.6418962998212026-09-25snapshot
Muse Spark 1.11475.61540111991142026-07-20snapshot
Claude Opus 4.7 Thinking1476.53845708086872026-07-30snapshot
Gemini 3.5 Flash1479.31788207159022026-09-01snapshot
GPT-5.41475.9431319490572026-08-03snapshot
Gemini 3.6 Flash1475.03337497458482026-07-21snapshot
Claude Opus 51479.2740006457892026-09-01snapshot
Claude Opus 51488.96966629876852026-10-08snapshot
Qwen3 81480.56729361543282026-08-12snapshot
Muse Spark 1.21482.63531240042812026-10-02snapshot
Claude Opus 4.61493.47999006415472026-10-08snapshot
Gemini 3.7 Flash1485.67681744116132026-09-13snapshot
Claude Fable 5.11498.4762533792962026-10-08snapshot
Gemini 3.8 Flash1489.9930800951432026-09-13snapshot
Muse Spark 1.31482.83365153980732026-09-13snapshot
Claude Fable 51483.78038624103972026-10-08snapshot
Mimo V2.6 Pro1483.84040559838012026-10-08snapshot
Gemini 4 Argon1511.64989033979282026-10-08snapshot
Claude Opus 5.51509.20332680589262026-10-08snapshot