General preference

Which model people prefer on prompts they brought themselves, aggregated over pairwise votes. This measures what users like, which tracks capability without being the same thing — length, formatting, and tone move it too, which is why the style-controlled rating is the one shown.

Ordered by how many of these benchmarks place a model in their top 10, then by its best placing. Each column keeps its own order; nothing here is averaged.

Model	In top 10	Arena — Text Overall (style control)
Claude Fable 5 (Opus 4.8 fallback) anthropic	1/1	#1 1508
Claude Opus 4.6 anthropic	1/1	#2 1504
Claude Opus 4.7 anthropic	1/1	#3 1502
Muse Spark 1.2 meta	1/1	#4 1498
Claude Opus 5 anthropic	1/1	#5 1493
Muse Spark 1.1 meta	1/1	#6 1491
Gemini 3.7 Flash google	1/1	#7 1490
Kimi K3 kimi	1/1	#8 1489
Muse Spark meta	1/1	#9 1488
GLM-5.3 zai	1/1	#10 1487

A dash means the source does not cover that model, not that it scored zero. Models with no registry entry are excluded here because they cannot be joined across sources — they are still shown, flagged, on their own source's board.