Skip to content

Benchmark

Arena ELO Text Factuality - French

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
82
Scores recorded
95
Sources
1
IQ weight
2.21

Score distribution

1,476.5
Mean
1,478.8
Median
28.3
Std dev
1,443.6
10th percentile
1,513.7
90th percentile
1,398.5
Min
1,398.51,546.9
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.21 for chat models, measured over 82 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Gemini 4 ArgonHigh· LMArena data1,546.9
2Claude Opus 4.6High· LMArena data1,530.1
3Claude Opus 5High· LMArena data1,527.3
4GPT 5.4High· LMArena data1,523
5Qwen3.8 Max· LMArena data1,517.9
6Claude Fable 5High· LMArena data1,516.7
7Muse Spark 1.3Max· LMArena data1,516.1
8GPT 5.5High· LMArena data1,515.4
9Claude Opus 4.7High· LMArena data1,514.5
10Claude Opus 5.5High· LMArena data1,513.7
11GPT 5.5· LMArena data1,511.1
12Gemini 3.6 FlashHigh· LMArena data1,509.4
13GLM 5.3Max· LMArena data1,508.4
14Gemini 3.5 FlashHigh· LMArena data1,508
15Claude Opus 4.6· LMArena data1,507.9
16GPT 5.4· LMArena data1,506.9
17GPT-5.6 TerraxHigh· LMArena data1,504.3
18Claude Opus 4.8High· LMArena data1,502.2
19Claude Opus 4.7· LMArena data1,501.2
20Claude Opus 4.8· LMArena data1,500.8
21MiMo V2.5 Pro· LMArena data1,500.4
22Gemini 3.1 Pro Preview· LMArena data1,499.6
23GPT-5.6 SolxHigh· LMArena data1,499
24Kimi K3Max· LMArena data1,498
25GLM 5.3 Flash· LMArena data1,497.8
26Claude Opus 5Max· LMArena data1,497.2
27Claude Opus 4.5High, 32K, 20251101· LMArena data1,496.2
28GPT-5.6 LunaxHigh· LMArena data1,495.8
29Grok 4.20Reasoning, 0309, beta· LMArena data1,492.4
30Kimi K2.6· LMArena data1,490.7
31Muse Spark 1.1· LMArena data1,490.6
32Gemini 3.8 FlashHigh· LMArena data1,490.3
33Gemini 2.5 Pro· LMArena data1,488.3
34GPT 5.2High· LMArena data1,487.6
35DeepSeek V4 Pro· LMArena data1,487.4
36Claude Sonnet 4.6· LMArena data1,487.1
37GPT-5.4 miniHigh· LMArena data1,485.3
38Claude Sonnet 5High· LMArena data1,484.8
39ChatGPT-4o Latest20250326· LMArena data1,484.1
40GLM 5.2Max· LMArena data1,482.4
41Grok 4.20 Multi-Agent0309, beta· LMArena data1,482
42Grok 4.7xHigh· LMArena data1,481.6
43ERNIE 5.1· LMArena data1,481.6
44Gemini 3.5 FlashMedium· LMArena data1,481.5
45GLM 5.1· LMArena data1,480.4
46Qwen3.7 Plus· LMArena data1,479.6
47GPT 5.1High· LMArena data1,479.2
48GPT 5.2· LMArena data1,478.8
49Inkling FP4· LMArena data1,478.4
50Claude Opus 4.520251101· LMArena data1,475.3

Showing 1–50 of 95 results.

Data credits: LMArena, CC BY 4.0 (licence)