---
title: "Text · factuality: model rankings and source evidence · Undominated.ai"
canonical: https://undominated.ai/rankings/arena-text-factuality/
description: "Text · factuality ranked by Arena score. Inspect exact model variants, source dates, measurement limits and downloadable evidence."
---

# Text · factuality: model rankings and source evidence · Undominated.ai

> Text · factuality ranked by Arena score. Inspect exact model variants, source dates, measurement limits and downloadable evidence.

[All rankings](/rankings/) / Language models

# Text · factuality

Human preference ratings for the overall category of this Arena task. Scores from different boards are not interchangeable.

**188 entries** · Arena score · Higher is better

Source publication: 2026-10-28. Retrieved 2026-11-05.

Source: [Arena (lmarena-ai)](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset) · [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/).

Ranks use confidence bounds. Several models can share a rank; the rows within a shared rank are displayed by score.

Text · factuality · Arena score

| Significance rank | Model / evaluated variant | Arena score | 95% confidence interval | Votes |
| --- | --- | --- | --- | --- |
| 1 | Gemini 4 Carbon (High) google | 1,521.59 | 1,513.37 – 1,529.8 | 4,403 |
| 1 | Claude Opus 5.5 (High) anthropic | 1,508.68 | 1,502.98 – 1,514.37 | 11,415 |
| 2 | Claude Fable 5.1 (Max) anthropic | 1,499.69 | 1,494.74 – 1,504.63 | 17,178 |
| 2 | Gemini 4 Argon (High) google | 1,499.54 | 1,494.09 – 1,504.99 | 11,889 |
| 3 | Muse Wonder (Max) meta | 1,494.33 | 1,487.98 – 1,500.68 | 7,749 |
| 3 | Claude Opus 4.6 (High) anthropic | 1,492.62 | 1,489.99 – 1,495.25 | 84,183 |
| 5 | Claude Opus 5 (Max) anthropic | 1,488.39 | 1,484.62 – 1,492.15 | 35,531 |
| 6 | Claude Opus 4.6 anthropic | 1,487.36 | 1,484.8 – 1,489.93 | 88,951 |
| 6 | GPT-5.5 (High) openai | 1,485.69 | 1,482.84 – 1,488.55 | 69,331 |
| 5 | MiMo-V2.6-Pro xiaomi | 1,484.5 | 1,476.52 – 1,492.48 | 4,037 |
| 7 | GPT-5.4 (High) openai | 1,484.26 | 1,481.31 – 1,487.2 | 64,664 |
| 7 | Claude Fable 5 (High) anthropic | 1,482.73 | 1,479.42 – 1,486.03 | 45,341 |
| 8 | GPT-5.5 openai | 1,481.83 | 1,479.01 – 1,484.65 | 70,892 |
| 8 | Gemini 3 Pro google | 1,481.24 | 1,477.83 – 1,484.65 | 41,121 |
| 6 | Qwen3.7-Max-Preview alibaba | 1,479.64 | 1,470.86 – 1,488.42 | 3,932 |
| 9 | Muse Spark 1.3 (Max) meta | 1,479.1 | 1,474.58 – 1,483.62 | 22,134 |
| 9 | Muse Spark 1.2 (xHigh) meta | 1,479.03 | 1,473.56 – 1,484.5 | 11,724 |
| 11 | Claude Opus 5 (High) anthropic | 1,478.25 | 1,475.35 – 1,481.14 | 73,167 |
| 11 | Claude Opus 4.7 (High) anthropic | 1,477.38 | 1,474.37 – 1,480.4 | 65,015 |
| 11 | Gemini 3.5 Flash (High) google | 1,477.16 | 1,473.83 – 1,480.48 | 48,376 |
| 11 | GPT-5.6 Sol (xHigh) openai | 1,476.4 | 1,473.07 – 1,479.73 | 43,796 |
| 11 | Gemini 3.7 Flash (High) google | 1,475.76 | 1,471.88 – 1,479.64 | 32,289 |
| 10 | Claude Sonnet 5.5 (xHigh) anthropic | 1,475.57 | 1,469.79 – 1,481.35 | 10,033 |
| 13 | Claude Opus 4.7 anthropic | 1,475.54 | 1,472.47 – 1,478.62 | 66,042 |
| 14 | GPT-5.4 openai | 1,474.85 | 1,471.95 – 1,477.75 | 67,730 |
| 13 | Qwen3.5-Max-Preview alibaba | 1,474.59 | 1,470.24 – 1,478.94 | 22,544 |
| 15 | Gemini 3.8 Flash (High) google | 1,472 | 1,468.24 – 1,475.76 | 47,803 |
| 15 | Gemini 3.6 Flash (High) google | 1,471.88 | 1,468.26 – 1,475.51 | 36,028 |
| 18 | Gemini 3.1 Pro Preview google | 1,471.68 | 1,469.42 – 1,473.94 | 132,208 |
| 16 | Kimi K3 (Max) moonshot | 1,471.51 | 1,468.01 – 1,475.01 | 36,780 |
| 16 | Qwen3.8-Max alibaba | 1,471.11 | 1,467.1 – 1,475.12 | 31,915 |
| 21 | Claude Opus 4.8 (High) anthropic | 1,469.81 | 1,466.73 – 1,472.9 | 63,939 |
| 18 | DeepSeek-V4.1-Flash-20260910 (Max) deepseek | 1,468.99 | 1,463.82 – 1,474.15 | 13,657 |
| 23 | GPT-6 Luna (Max) openai | 1,468.71 | 1,465.48 – 1,471.93 | 86,478 |
| 23 | GPT-5.6 Terra (xHigh) openai | 1,468.68 | 1,465.47 – 1,471.89 | 48,166 |
| 21 | Muse Spark meta | 1,467.63 | 1,462.2 – 1,473.07 | 14,122 |
| 24 | Gemini 3 Flash google | 1,467.5 | 1,463.55 – 1,471.45 | 30,651 |
| 25 | Gemini 3.5 Flash (Medium) google | 1,467.49 | 1,464.15 – 1,470.83 | 47,196 |
| 24 | GPT-6 Astra (Max) openai | 1,466.52 | 1,461.51 – 1,471.53 | 15,038 |
| 25 | GLM-5.3 (Max) zai | 1,466.11 | 1,461.93 – 1,470.29 | 26,606 |
| 24 | GPT-6.1 Sol (Max) openai | 1,465.56 | 1,459.56 – 1,471.55 | 9,367 |
| 28 | Grok 4.5 xai | 1,465.19 | 1,461.75 – 1,468.64 | 39,801 |
| 31 | Claude Opus 4.8 anthropic | 1,464.73 | 1,461.7 – 1,467.76 | 65,075 |
| 31 | MiMo-V2.5-Pro xiaomi | 1,464.33 | 1,461.53 – 1,467.12 | 77,666 |
| 31 | GLM-5.3-Flash zai | 1,463.79 | 1,460.01 – 1,467.57 | 41,642 |
| 33 | GLM-5.2 (Max) zai | 1,462.6 | 1,459.37 – 1,465.83 | 53,702 |
| 33 | Muse Spark 1.1 meta | 1,462.44 | 1,458.96 – 1,465.91 | 46,354 |
| 35 | Gemini 2.5 Pro google | 1,461.74 | 1,459.1 – 1,464.39 | 69,844 |
| 38 | Claude Sonnet 4.6 anthropic | 1,460.48 | 1,457.6 – 1,463.36 | 70,653 |
| 36 | GPT-5.6 Luna (xHigh) openai | 1,460.47 | 1,457.07 – 1,463.87 | 41,317 |
| 40 | Claude Opus 4.5 (20251101) anthropic | 1,459.11 | 1,456.38 – 1,461.84 | 71,687 |
| 38 | Claude Opus 4.5 (High, 32K, 20251101) anthropic | 1,458.92 | 1,455.45 – 1,462.39 | 36,607 |
| 44 | GPT-5.1 (High) openai | 1,458.1 | 1,454.79 – 1,461.41 | 40,765 |
| 44 | Grok 4.20 Multi-Agent Beta (0309) xai | 1,457.01 | 1,453.99 – 1,460.04 | 64,827 |
| 45 | Kimi K2.6 moonshot | 1,456.23 | 1,452.56 – 1,459.89 | 39,929 |
| 45 | ERNIE 5.1 baidu | 1,455.99 | 1,452.15 – 1,459.82 | 39,414 |
| 49 | GLM-5.1 zai | 1,455.93 | 1,452.98 – 1,458.88 | 66,633 |
| 49 | Grok 4.20 Beta (Reasoning, 0309) xai | 1,455.66 | 1,452.61 – 1,458.7 | 66,343 |
| 49 | DeepSeek-V4-Pro deepseek | 1,454.65 | 1,451.52 – 1,457.78 | 57,532 |
| 46 | DeepSeek-V4-Pro (High, 20260813) deepseek | 1,453.99 | 1,448.49 – 1,459.48 | 9,787 |
| 52 | Qwen3.7-Plus alibaba | 1,452.41 | 1,448.92 – 1,455.9 | 42,011 |
| 51 | Grok 4.20 Beta 1 xai | 1,452.37 | 1,448.25 – 1,456.49 | 27,693 |
| 45 | Qwen3.6-Max-Preview alibaba | 1,452.35 | 1,444.76 – 1,459.95 | 5,419 |
| 53 | Claude Sonnet 5 (High) anthropic | 1,452.17 | 1,448.92 – 1,455.41 | 52,952 |
| 53 | Inkling thinky | 1,451.72 | 1,448.34 – 1,455.1 | 45,472 |
| 44 | amazon-nova-experimental-chat (26-02-10) amazon | 1,451.6 | 1,442.8 – 1,460.4 | 3,523 |
| 53 | GLM-5 zai | 1,451.53 | 1,447.69 – 1,455.37 | 29,021 |
| 55 | Gemini 3 Flash (Minimal) google | 1,451.15 | 1,448.62 – 1,453.69 | 90,228 |
| 52 | MiMo-V2.6-Flash xiaomi | 1,450.73 | 1,445.55 – 1,455.9 | 15,312 |
| 54 | GPT-5.2 Chat (20260210) openai | 1,450.5 | 1,446.89 – 1,454.11 | 35,651 |
| 55 | Grok 4.6 (High) xai | 1,450.09 | 1,446.36 – 1,453.81 | 33,824 |
| 52 | Gemma 4 31B google | 1,449.34 | 1,442.47 – 1,456.2 | 6,135 |
| 60 | Claude Sonnet 4.5 (20250929) anthropic | 1,448.52 | 1,445.97 – 1,451.07 | 77,498 |
| 60 | Gemini 3.5 Flash-Lite google | 1,447.58 | 1,443.87 – 1,451.29 | 35,740 |
| 60 | DeepSeek-V4-Pro Preview (High) deepseek | 1,447.42 | 1,444.23 – 1,450.61 | 54,811 |
| 56 | Step 5 Preview stepfun | 1,447.16 | 1,441.62 – 1,452.7 | 11,355 |
| 60 | GLM-4.6 zai | 1,446.83 | 1,442.95 – 1,450.71 | 28,684 |
| 50 | DeepSeek-V3.2-Exp (Thinking) deepseek | 1,446.45 | 1,435.75 – 1,457.15 | 2,523 |
| 60 | GPT-5.4 mini (High) openai | 1,446.3 | 1,443.32 – 1,449.29 | 63,316 |
| 62 | Claude Sonnet 4.5 (High, 32K, 20250929) anthropic | 1,446.23 | 1,443.66 – 1,448.8 | 74,727 |
| 60 | Hy3 tencent | 1,445.74 | 1,440.74 – 1,450.73 | 13,840 |
| 62 | ERNIE 5.0 (0110) baidu | 1,445.22 | 1,441.67 – 1,448.77 | 35,909 |
| 67 | Kimi K2.5 (Thinking) moonshot | 1,444.84 | 1,442.11 – 1,447.58 | 74,317 |
| 60 | Gemma 4 26B A4B google | 1,444.54 | 1,437.65 – 1,451.43 | 6,088 |
| 68 | Qwen3.5-397B-A17B alibaba | 1,444.39 | 1,441.92 – 1,446.86 | 97,351 |
| 67 | GPT-5.1 openai | 1,443.93 | 1,440.67 – 1,447.19 | 43,458 |
| 60 | GLM-4.7 zai | 1,443.87 | 1,438.2 – 1,449.54 | 11,796 |
| 67 | ChatGPT-4o Latest (20250326) openai | 1,443.74 | 1,440.35 – 1,447.13 | 40,465 |
| 67 | MiMo-V2-Pro xiaomi | 1,443.24 | 1,439.05 – 1,447.43 | 25,385 |
| 67 | GPT-5.5 Instant openai | 1,442.76 | 1,438.59 – 1,446.94 | 27,128 |
| 66 | ERNIE-5.0-Preview (1203) baidu | 1,441.93 | 1,435.77 – 1,448.1 | 9,545 |
| 73 | GPT-5.2 openai | 1,441.57 | 1,438.99 – 1,444.14 | 83,183 |
| 66 | Qwen3-Max-Preview alibaba | 1,441.37 | 1,434.64 – 1,448.11 | 7,567 |
| 73 | GPT-5.2 (High) openai | 1,440.91 | 1,437.81 – 1,444.01 | 49,087 |
| 62 | Beam 1.0 501B A23B reflection | 1,440.85 | 1,432.96 – 1,448.74 | 4,233 |
| 68 | GLM-5V-Turbo zai | 1,440.78 | 1,435.11 – 1,446.45 | 10,476 |
| 71 | GPT-6 Sol (Max) openai | 1,439.79 | 1,434.75 – 1,444.82 | 15,807 |
| 66 | ERNIE-5.0-Preview (1022) baidu | 1,439.54 | 1,430.87 – 1,448.21 | 4,017 |
| 79 | MiMo-V2.5 xiaomi | 1,438.81 | 1,435.37 – 1,442.25 | 47,279 |
| 80 | DeepSeek-V4-Flash deepseek | 1,438.78 | 1,435.47 – 1,442.08 | 51,876 |
| 83 | Qwen3.6-Plus alibaba | 1,437.44 | 1,434 – 1,440.88 | 48,068 |
| 86 | DeepSeek-V3.2 deepseek | 1,436.54 | 1,433.36 – 1,439.71 | 47,123 |
| 85 | Qwen3.8-27B alibaba | 1,436.14 | 1,431.92 – 1,440.35 | 27,779 |
| 88 | Gemini 2.5 Flash google | 1,436.03 | 1,433.46 – 1,438.59 | 70,533 |
| 81 | DeepSeek-V3.2-Exp deepseek | 1,435.99 | 1,430.08 – 1,441.89 | 11,902 |
| 86 | MiMo-V2-Omni xiaomi | 1,435.29 | 1,430.55 – 1,440.02 | 20,572 |
| 89 | Claude Opus 4.1 (20250805) anthropic | 1,434.92 | 1,431.6 – 1,438.25 | 41,534 |
| 86 | Grok 4.7 (xHigh) xai | 1,434.49 | 1,429.57 – 1,439.42 | 16,442 |
| 92 | Mistral Large 3 mistral | 1,434.14 | 1,431.72 – 1,436.55 | 87,371 |
| 92 | Grok 4.1 (Thinking) xai | 1,434.04 | 1,431.12 – 1,436.97 | 65,534 |
| 92 | MiniMax-M3 minimax | 1,433.99 | 1,430.98 – 1,437 | 68,611 |
| 92 | DeepSeek-V4-Flash Preview (High) deepseek | 1,433.54 | 1,430.12 – 1,436.96 | 51,698 |
| 88 | Mistral Large 4 mistral | 1,433.31 | 1,427.87 – 1,438.74 | 12,066 |
| 92 | Claude Opus 4.1 (Thinking 16K, 20250805) anthropic | 1,433.16 | 1,429.09 – 1,437.23 | 25,935 |
| 92 | Gemini 3.1 Flash-Lite Preview google | 1,432.81 | 1,429.76 – 1,435.87 | 63,927 |
| 95 | DeepSeek-V3.2 (Thinking) deepseek | 1,432.18 | 1,428.93 – 1,435.43 | 41,144 |
| 91 | NVIDIA-Nemotron-3-Ultra-550B-A55B (NVFP4) nvidia | 1,431.6 | 1,425.48 – 1,437.73 | 11,563 |
| 100 | Grok 4.1 xai | 1,431 | 1,428.13 – 1,433.87 | 67,654 |
| 101 | Qwen3-235B-A22B-Instruct (2507) alibaba | 1,430.76 | 1,428.1 – 1,433.42 | 66,630 |
| 92 | Mistral Medium 3.5 mistral | 1,430 | 1,424.2 – 1,435.8 | 11,729 |
| 97 | Kimi K2.5 (Instant) moonshot | 1,428.82 | 1,422.78 – 1,434.86 | 8,250 |
| 103 | LongCat-Flash-Chat Exp (2602) meituan | 1,428.69 | 1,424.62 – 1,432.76 | 29,163 |
| 106 | Mistral Medium 3.1 (2508) mistral | 1,428.69 | 1,426.02 – 1,431.35 | 68,688 |
| 106 | Qwen3.5-122B-A10B alibaba | 1,427.61 | 1,423.76 – 1,431.46 | 29,517 |
| 113 | Grok 4.3 xai | 1,426.79 | 1,423.92 – 1,429.67 | 71,249 |
| 114 | Inkling-Small thinky | 1,425.65 | 1,421.9 – 1,429.39 | 36,999 |
| 116 | Gemini 2.5 Flash Preview (09-2025) google | 1,424.17 | 1,419.84 – 1,428.51 | 23,417 |
| 104 | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4) nvidia | 1,424.02 | 1,416.29 – 1,431.75 | 4,178 |
| 103 | amazon-nova-experimental-chat (12-10) amazon | 1,423.76 | 1,414.56 – 1,432.96 | 3,549 |
| 120 | Kimi K2 Thinking Turbo moonshot | 1,422.92 | 1,420.12 – 1,425.72 | 62,351 |
| 121 | MiniMax-M2.7 minimax | 1,422.22 | 1,419.54 – 1,424.9 | 89,123 |
| 121 | MiMo-V2-Flash (Non-Thinking) xiaomi | 1,421.7 | 1,418.55 – 1,424.86 | 47,349 |
| 116 | GPT-5 Chat openai | 1,421.6 | 1,414.99 – 1,428.21 | 7,764 |
| 119 | GPT-5 (High) openai | 1,420.42 | 1,413.6 – 1,427.23 | 6,923 |
| 125 | Dola Seed 2.0 Pro bytedance | 1,419.86 | 1,416.86 – 1,422.86 | 78,771 |
| 125 | Qwen3.5-27B alibaba | 1,419.34 | 1,415.46 – 1,423.23 | 28,461 |
| 120 | Claude Haiku 5.5 (High) anthropic | 1,419.08 | 1,412.68 – 1,425.48 | 7,890 |
| 127 | Claude Haiku 4.5 (20251001) anthropic | 1,418.46 | 1,416.43 – 1,420.48 | 147,416 |
| 121 | Qwen3-Next-80B-A3B-Instruct alibaba | 1,418.3 | 1,411.63 – 1,424.96 | 7,740 |
| 123 | Grok 4 (0709) xai | 1,418.06 | 1,412.14 – 1,423.97 | 10,438 |
| 127 | GPT-5.3 Chat openai | 1,417.45 | 1,413.7 – 1,421.2 | 34,113 |
| 125 | Grok 4 Fast (Reasoning) xai | 1,417.05 | 1,411.1 – 1,423 | 10,657 |
| 129 | GPT-5.4 nano (High) openai | 1,416.81 | 1,413.8 – 1,419.81 | 62,208 |
| 130 | Step-3.5-Flash stepfun | 1,416.07 | 1,413.07 – 1,419.06 | 59,956 |
| 130 | amazon-nova-experimental-chat (11-10) amazon | 1,414.85 | 1,410.86 – 1,418.83 | 25,511 |
| 103 | Qwen3-VL-235B-A22B-Instruct alibaba | 1,414.34 | 1,396.72 – 1,431.96 | 3,040 |
| 130 | Solar Pro 4 upstage | 1,414.32 | 1,410.02 – 1,418.61 | 26,159 |
| 132 | Qwen3.5-Flash alibaba | 1,413.61 | 1,410.52 – 1,416.7 | 61,569 |
| 133 | Grok 4.1 Fast (Reasoning) xai | 1,413.38 | 1,410.41 – 1,416.34 | 56,561 |
| 130 | Hy3 preview tencent | 1,411.97 | 1,405.08 – 1,418.87 | 6,866 |
| 135 | MiniMax-M2.1 Preview minimax | 1,410.53 | 1,405.74 – 1,415.33 | 16,776 |
| 136 | MiMo-V2-Flash (Thinking) xiaomi | 1,409.18 | 1,403.46 – 1,414.89 | 11,049 |
| 137 | GPT-4.1 (2025-04-14) openai | 1,407.48 | 1,400.9 – 1,414.06 | 7,877 |
| 145 | Qwen3.5-35B-A3B alibaba | 1,407.08 | 1,403.25 – 1,410.91 | 30,222 |
| 149 | Gemini 2.5 Flash-Lite Preview (Non-Thinking, 09-2025) google | 1,406.4 | 1,402.93 – 1,409.87 | 37,558 |
| 146 | o3 (2025-04-16) openai | 1,404.14 | 1,397.51 – 1,410.76 | 7,558 |
| 132 | Kimi K2 Preview (0905) moonshot | 1,402.3 | 1,387.92 – 1,416.67 | 1,453 |
| 144 | amazon-nova-experimental-chat (26-01-10) amazon | 1,402.26 | 1,393.42 – 1,411.1 | 3,534 |
| 148 | Muse Glimmer meta | 1,401.84 | 1,393.46 – 1,410.22 | 3,884 |
| 131 | Qwen3-Coder-480B-A35B-Instruct alibaba | 1,399.8 | 1,382.35 – 1,417.25 | 3,444 |
| 150 | amazon-nova-experimental-chat (10-20) amazon | 1,399.27 | 1,393.4 – 1,405.13 | 11,198 |
| 154 | GLM-4.5-Air zai | 1,394.11 | 1,387.31 – 1,400.9 | 7,272 |
| 155 | GPT-5 mini (High) openai | 1,393.43 | 1,386.55 – 1,400.31 | 6,888 |
| 154 | GLM-4.6V zai | 1,391.05 | 1,380.43 – 1,401.67 | 2,515 |
| 157 | NVIDIA-Nemotron-3-Super-120B-A12B nvidia | 1,389.18 | 1,382.28 – 1,396.09 | 7,843 |
| 160 | MiniMax-M2.5 minimax | 1,385.01 | 1,381.51 – 1,388.5 | 42,435 |
| 160 | gpt-oss-120b openai | 1,384.48 | 1,377.84 – 1,391.13 | 7,393 |
| 163 | Amazon Nova 2 Lite amazon | 1,379.31 | 1,373.58 – 1,385.05 | 11,939 |
| 160 | Gemma 3 27B IT google | 1,379.13 | 1,367.99 – 1,390.28 | 2,384 |
| 163 | Granite 4.2 30B ibm | 1,375.4 | 1,366.31 – 1,384.49 | 3,389 |
| 166 | INTELLECT-3 Organization not supplied | 1,372.64 | 1,364.83 – 1,380.46 | 5,111 |
| 168 | Trinity Large Preview Organization not supplied | 1,371.47 | 1,367.62 – 1,375.31 | 30,952 |
| 168 | Trinity Large Thinking Organization not supplied | 1,370.25 | 1,366.04 – 1,374.46 | 30,131 |
| 168 | NVIDIA-Nemotron-3.5-Lightning-30B-A3B (NVFP4) nvidia | 1,369.53 | 1,364.86 – 1,374.21 | 24,230 |
| 167 | MiniMax-M1 minimax | 1,369.44 | 1,358.62 – 1,380.26 | 2,569 |
| 168 | GLM-4.7-Flash zai | 1,369.4 | 1,363.87 – 1,374.93 | 11,879 |
| 168 | Mercury 2 inception-ai | 1,367.7 | 1,358.2 – 1,377.19 | 3,190 |
| 168 | o4-mini (2025-04-16) openai | 1,367.48 | 1,357.21 – 1,377.74 | 2,721 |
| 169 | Command A (03-2025) cohere | 1,362.81 | 1,356.34 – 1,369.28 | 7,703 |
| 171 | NVIDIA-Nemotron-3-Nano-30B-A3B (BF16) nvidia | 1,362.2 | 1,356.95 – 1,367.44 | 15,454 |
| 171 | MiniMax-M2 minimax | 1,359.73 | 1,352.51 – 1,366.95 | 6,683 |
| 182 | Granite 4.2 8B ibm | 1,342.05 | 1,332.29 – 1,351.81 | 3,251 |
| 182 | Olmo 3.1 32B Instruct allenai | 1,341.44 | 1,335.61 – 1,347.28 | 11,693 |
| 183 | Granite 4.2 3B ibm | 1,323.88 | 1,313.95 – 1,333.81 | 3,249 |
| 184 | Olmo 3 32B Think allenai | 1,318.64 | 1,310.84 – 1,326.45 | 5,704 |
| 184 | Granite 4.1 8B ibm | 1,310.35 | 1,301.11 – 1,319.58 | 4,177 |
| 184 | Mercury inception-ai | 1,304.24 | 1,291.68 – 1,316.8 | 1,937 |
| 186 | Olmo 3.1 32B Think allenai | 1,299.9 | 1,292.97 – 1,306.83 | 8,099 |

## What this ranking can tell you

 - The population is the models in this source board, including variants and models without an available API price.
- Ranks use the published confidence bounds: one plus the number of models whose lower bound exceeds this model’s upper bound. Shared ranks are not proof of equal performance.

## Check the source

The original evidence is retained byte for byte. The downloadable board includes the source URLs, revision and SHA-256 hashes.

[Download this board and provenance →](/data/rankings/arena-text-factuality.json)

 Source files and licences - [metadata · a7a159f557a1](/data/rankings/evidence/a7a159f557a1e594a5d41218fe8d53c63c63daa39b655e51d6fc5b482a17d059.json) · [upstream source ↗](https://huggingface.co/api/datasets/lmarena-ai/leaderboard-dataset)
- [licence · 88a46d472f35](/data/rankings/evidence/88a46d472f35274fb6b7a852c3d16c8ee8007d51b37f290161537bec26895ee4.txt) · [upstream source ↗](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/raw/8b6deb69ace01ee8dcda9e46eba7a5e8a808ec53/README.md)
- [rows · 26b14382e6b8](/data/rankings/evidence/26b14382e6b8561b324cba0c156b13f3b74cae7abf321327a609c407331f9d25.parquet) · [upstream source ↗](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/8b6deb69ace01ee8dcda9e46eba7a5e8a808ec53/text_factuality/latest-00000-of-00001.parquet)

[Explore other model types →](/rankings/)[Read the benchmark guide →](/benchmarks/)

## Continue your investigation

 - [Understand the measurements](/benchmarks/)
- [Compare task performance](/best/)
- [Compare embedding models](/embeddings/)
- [Inspect ranking rules](/methodology/)