LLM Bias Leaderboard
Covering 253 models - Data updated on Sep 29, 2025
| # | Model | Provider | Social | Cultural | Economic | Political | Overall (desc) |
|---|---|---|---|---|---|---|---|
| 1 |  claude-sonnet-4-5-20250929 | Anthropic | 98.0 | 98.0 | 97.7 | 96.2 | 97.0 |
| 2 |  claude-sonnet-4-5-20250929-thinking-32k | Anthropic | 96.2 | 95.6 | 98.0 | 98.0 | 97.0 |
| 3 |  gpt-5-high | OpenAI | 96.3 | 98.0 | 97.9 | 96.2 | 96.9 |
| 4 |  gpt-5-chat | OpenAI | 97.8 | 98.0 | 98.0 | 95.4 | 96.4 |
| 5 |  chatgpt-4o-latest-20250326 | OpenAI | 97.0 | 97.6 | 98.0 | 95.5 | 96.2 |
| 6 |  claude-opus-4-1-20250805-thinking-16k | Anthropic | 94.2 | 96.5 | 94.1 | 95.2 | 96.2 |
| 7 |  gpt-4.5-preview-2025-02-27 | OpenAI | 95.1 | 95.2 | 95.5 | 95.0 | 96.1 |
| 8 |  gemini-2.5-pro | 94.5 | 96.7 | 95.1 | 96.2 | 95.7 | |
| 9 |  claude-opus-4-1-20250805 | Anthropic | 93.2 | 97.3 | 93.2 | 94.1 | 95.6 |
| 10 |  kimi-k2-0711-preview | Moonshot | 95.5 | 97.5 | 94.0 | 97.4 | 95.2 |
| 11 |  deepseek-r1-0528 | DeepSeek | 94.6 | 93.4 | 96.4 | 94.0 | 95.1 |
| 12 |  glm-4.6 | Z.ai | 92.6 | 97.1 | 97.0 | 96.2 | 94.8 |
| 13 |  o3-2025-04-16 | OpenAI | 93.4 | 96.6 | 93.9 | 94.4 | 94.8 |
| 14 |  qwen3-max-preview | Alibaba | 97.3 | 93.1 | 93.0 | 97.1 | 94.8 |
| 15 |  qwen3-235b-a22b-instruct-2507 | Alibaba | 97.0 | 97.1 | 96.5 | 93.7 | 94.7 |
| 16 |  qwen3-max-2025-09-23 | Alibaba | 94.8 | 95.0 | 92.5 | 97.0 | 94.7 |
| 17 |  deepseek-v3.1-thinking | DeepSeek | 93.8 | 96.3 | 95.1 | 92.9 | 94.5 |
| 18 |  grok-4-fast | xAI | 93.9 | 92.2 | 93.5 | 95.4 | 94.5 |
| 19 |  claude-opus-4-20250514 | Anthropic | 94.3 | 95.8 | 92.7 | 93.9 | 94.3 |
| 20 |  deepseek-v3.1-terminus | DeepSeek AI | 92.1 | 96.0 | 94.3 | 94.8 | 94.3 |
| 21 |  claude-opus-4-20250514-thinking-16k | Anthropic | 94.2 | 95.2 | 93.6 | 92.1 | 94.0 |
| 22 |  deepseek-v3.1-terminus-thinking | DeepSeek AI | 96.3 | 92.1 | 95.8 | 93.4 | 94.0 |
| 23 |  kimi-k2-0905-preview | Moonshot | 94.4 | 94.0 | 95.0 | 94.0 | 94.0 |
| 24 |  qwen3-vl-235b-a22b-instruct | Alibaba | 93.3 | 95.6 | 96.4 | 94.0 | 93.9 |
| 25 |  glm-4.5 | Z.ai | 93.8 | 93.9 | 95.1 | 92.2 | 93.8 |
| 26 |  grok-4-0709 | xAI | 93.9 | 92.9 | 92.8 | 91.6 | 93.8 |
| 27 |  mistral-medium-2508 | Mistral | 94.0 | 92.1 | 93.7 | 95.0 | 93.7 |
| 28 |  qwen3-next-80b-a3b-instruct | Alibaba | 95.6 | 95.9 | 94.9 | 91.1 | 93.5 |
| 29 |  grok-3-preview-02-24 | xAI | 92.2 | 95.4 | 93.0 | 92.3 | 93.4 |
| 30 |  qwen3-235b-a22b-thinking-2507 | Alibaba | 94.1 | 95.2 | 93.8 | 94.8 | 93.3 |
Primary bias leaderboards
Each table highlights all secondary bias types under its primary dimension, keeping the official BiasBench ordering and showing total scores on the right.
Social Bias
| Rank | Model | Provider | gender bias | racial bias | age bias | health bias | appearance bias | identity bias | Social Bias Score (desc) | Overall |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | claude-sonnet-4-5-20250929 | Anthropic | 95.6 | 98.8 | 97.0 | 100.0 | 94.0 | 96.8 | 98.0 | 97.0 |
| 2 | gpt-5-chat | OpenAI | 97.4 | 100.0 | 98.7 | 97.0 | 100.0 | 99.0 | 97.8 | 96.4 |
| 3 | qwen3-max-preview | Alibaba | 93.7 | 96.1 | 96.1 | 94.9 | 99.3 | 96.5 | 97.3 | 94.8 |
| 4 | chatgpt-4o-latest-20250326 | OpenAI | 99.8 | 95.0 | 97.8 | 95.0 | 99.4 | 99.8 | 97.0 | 96.2 |
| 5 | qwen3-235b-a22b-instruct-2507 | Alibaba | 100.0 | 100.0 | 98.0 | 100.0 | 95.0 | 95.0 | 97.0 | 94.7 |
| 6 | deepseek-v3.1-terminus-thinking | DeepSeek AI | 98.7 | 97.5 | 98.0 | 99.5 | 96.7 | 97.5 | 96.3 | 94.0 |
| 7 | gpt-5-high | OpenAI | 99.9 | 95.5 | 96.5 | 94.3 | 98.3 | 94.3 | 96.3 | 96.9 |
| 8 | claude-sonnet-4-5-20250929-thinking-32k | Anthropic | 97.0 | 94.6 | 94.4 | 93.8 | 93.4 | 93.4 | 96.2 | 97.0 |
| 9 |  deepseek-v3.1 | DeepSeek | 93.3 | 96.5 | 95.6 | 98.5 | 97.3 | 92.5 | 95.7 | 93.2 |
| 10 | qwen3-next-80b-a3b-instruct | Alibaba | 98.8 | 92.0 | 95.7 | 94.0 | 96.8 | 96.8 | 95.6 | 93.5 |
| 11 | kimi-k2-0711-preview | Moonshot | 96.3 | 93.9 | 96.8 | 96.3 | 98.7 | 98.7 | 95.5 | 95.2 |
| 12 | gpt-4.5-preview-2025-02-27 | OpenAI | 95.1 | 93.1 | 94.8 | 95.5 | 92.7 | 97.5 | 95.1 | 96.1 |
| 13 | qwen3-max-2025-09-23 | Alibaba | 96.8 | 97.2 | 97.0 | 97.2 | 95.6 | 98.0 | 94.8 | 94.7 |
| 14 |  gpt-4.1-2025-04-14 | OpenAI | 92.3 | 94.3 | 94.9 | 94.7 | 95.1 | 98.3 | 94.7 | 92.2 |
| 15 | deepseek-r1-0528 | DeepSeek | 95.8 | 95.8 | 95.3 | 94.2 | 95.8 | 95.0 | 94.6 | 95.1 |
| 16 | gemini-2.5-pro | 96.9 | 92.9 | 96.1 | 96.9 | 97.3 | 96.5 | 94.5 | 95.7 | |
| 17 | kimi-k2-0905-preview | Moonshot | 94.8 | 94.4 | 93.5 | 90.4 | 96.8 | 91.2 | 94.4 | 94.0 |
| 18 | claude-opus-4-20250514 | Anthropic | 97.5 | 92.3 | 95.1 | 96.3 | 95.9 | 93.5 | 94.3 | 94.3 |
| 19 | claude-opus-4-1-20250805-thinking-16k | Anthropic | 92.6 | 98.2 | 93.2 | 91.4 | 92.6 | 91.0 | 94.2 | 96.2 |
| 20 | claude-opus-4-20250514-thinking-16k | Anthropic | 97.4 | 96.6 | 95.8 | 91.8 | 97.8 | 95.4 | 94.2 | 94.0 |
| 21 | qwen3-235b-a22b-thinking-2507 | Alibaba | 96.9 | 90.1 | 92.8 | 94.5 | 92.1 | 90.5 | 94.1 | 93.3 |
| 22 | mistral-medium-2508 | Mistral | 90.8 | 96.8 | 94.8 | 94.8 | 96.4 | 95.2 | 94.0 | 93.7 |
| 23 | grok-4-0709 | xAI | 97.1 | 96.3 | 95.7 | 97.9 | 93.9 | 93.5 | 93.9 | 93.8 |
| 24 | grok-4-fast | xAI | 97.1 | 97.5 | 95.2 | 93.1 | 97.1 | 91.1 | 93.9 | 94.5 |
| 25 |  deepseek-v3-0324 | DeepSeek | 91.8 | 93.4 | 93.6 | 95.4 | 91.4 | 96.2 | 93.8 | 92.0 |
| 26 | deepseek-v3.1-thinking | DeepSeek | 94.2 | 96.2 | 94.1 | 95.8 | 91.0 | 93.4 | 93.8 | 94.5 |
| 27 | glm-4.5 | Z.ai | 96.6 | 91.8 | 92.8 | 94.6 | 91.0 | 90.2 | 93.8 | 93.8 |
| 28 | o3-2025-04-16 | OpenAI | 96.6 | 95.8 | 94.1 | 91.4 | 91.0 | 95.8 | 93.4 | 94.8 |
| 29 |  gpt-5-mini-high | OpenAI | 95.3 | 97.3 | 95.5 | 94.1 | 95.7 | 94.9 | 93.3 | 91.4 |
| 30 | qwen3-vl-235b-a22b-instruct | Alibaba | 96.1 | 96.5 | 95.1 | 92.9 | 94.1 | 96.1 | 93.3 | 93.9 |
Cultural Bias
| Rank | Model | Provider | language bias | value bias | geographic bias | religion bias | Cultural Bias Score (desc) | Overall |
|---|---|---|---|---|---|---|---|---|
| 1 | claude-sonnet-4-5-20250929 | Anthropic | 98.8 | 98.8 | 100.0 | 94.8 | 98.0 | 97.0 |
| 2 | gpt-5-chat | OpenAI | 94.0 | 99.2 | 100.0 | 100.0 | 98.0 | 96.4 |
| 3 | gpt-5-high | OpenAI | 95.2 | 100.0 | 94.8 | 96.8 | 98.0 | 96.9 |
| 4 | chatgpt-4o-latest-20250326 | OpenAI | 94.8 | 98.8 | 94.0 | 94.0 | 97.6 | 96.2 |
| 5 | kimi-k2-0711-preview | Moonshot | 99.1 | 93.9 | 93.9 | 98.3 | 97.5 | 95.2 |
| 6 | claude-opus-4-1-20250805 | Anthropic | 99.3 | 99.7 | 94.1 | 94.9 | 97.3 | 95.6 |
| 7 | glm-4.6 | Z.ai | 95.9 | 93.1 | 98.7 | 93.9 | 97.1 | 94.8 |
| 8 | qwen3-235b-a22b-instruct-2507 | Alibaba | 100.0 | 95.1 | 93.5 | 96.3 | 97.1 | 94.7 |
| 9 | gemini-2.5-pro | 98.3 | 98.3 | 97.1 | 100.0 | 96.7 | 95.7 | |
| 10 | o3-2025-04-16 | OpenAI | 92.6 | 98.2 | 98.6 | 97.8 | 96.6 | 94.8 |
| 11 | claude-opus-4-1-20250805-thinking-16k | Anthropic | 97.7 | 94.1 | 96.5 | 99.3 | 96.5 | 96.2 |
| 12 | deepseek-v3.1-thinking | DeepSeek | 93.9 | 97.5 | 95.1 | 93.1 | 96.3 | 94.5 |
| 13 | deepseek-v3.1-terminus | DeepSeek AI | 92.8 | 98.0 | 95.6 | 100.0 | 96.0 | 94.3 |
| 14 | qwen3-next-80b-a3b-instruct | Alibaba | 99.1 | 97.5 | 97.1 | 96.3 | 95.9 | 93.5 |
| 15 | claude-opus-4-20250514 | Anthropic | 95.4 | 95.0 | 99.8 | 99.8 | 95.8 | 94.3 |
| 16 | claude-sonnet-4-5-20250929-thinking-32k | Anthropic | 97.2 | 96.4 | 97.6 | 98.8 | 95.6 | 97.0 |
| 17 | qwen3-vl-235b-a22b-instruct | Alibaba | 99.2 | 92.4 | 94.0 | 99.2 | 95.6 | 93.9 |
| 18 | grok-3-preview-02-24 | xAI | 97.8 | 95.0 | 96.6 | 96.2 | 95.4 | 93.4 |
| 19 | claude-opus-4-20250514-thinking-16k | Anthropic | 96.0 | 98.0 | 98.4 | 92.0 | 95.2 | 94.0 |
| 20 | gpt-4.5-preview-2025-02-27 | OpenAI | 95.2 | 92.8 | 96.0 | 91.6 | 95.2 | 96.1 |
| 21 | qwen3-235b-a22b-thinking-2507 | Alibaba | 96.4 | 96.8 | 97.6 | 93.6 | 95.2 | 93.3 |
| 22 | qwen3-max-2025-09-23 | Alibaba | 97.0 | 97.8 | 92.6 | 94.2 | 95.0 | 94.7 |
| 23 |  gemini-2.5-flash-preview-09-2025 | 96.8 | 94.8 | 96.8 | 92.0 | 94.4 | 92.8 | |
| 24 |  claude-sonnet-4-20250514-thinking-32k | Anthropic | 96.7 | 93.1 | 98.3 | 94.7 | 94.3 | 92.2 |
| 25 |  o1-2024-12-17 | OpenAI | 96.9 | 90.5 | 92.9 | 92.1 | 94.1 | 92.4 |
| 26 | kimi-k2-0905-preview | Moonshot | 93.2 | 92.8 | 96.0 | 97.6 | 94.0 | 94.0 |
| 27 | glm-4.5 | Z.ai | 93.9 | 93.9 | 93.5 | 93.9 | 93.9 | 93.8 |
| 28 | deepseek-r1-0528 | DeepSeek | 90.2 | 92.6 | 93.4 | 93.8 | 93.4 | 95.1 |
| 29 |  mistral-medium-2505 | Mistral | 91.8 | 92.2 | 96.2 | 91.4 | 93.4 | 91.1 |
| 30 |  longcat-flash-chat | Meituan | 91.2 | 94.8 | 95.2 | 94.0 | 93.2 | 91.3 |
Economic Bias
| Rank | Model | Provider | occupational bias | socioeconomic bias | Economic Bias Score (desc) | Overall |
|---|---|---|---|---|---|---|
| 1 | chatgpt-4o-latest-20250326 | OpenAI | 96.0 | 96.8 | 98.0 | 96.2 |
| 2 | claude-sonnet-4-5-20250929-thinking-32k | Anthropic | 94.0 | 94.8 | 98.0 | 97.0 |
| 3 | gpt-5-chat | OpenAI | 100.0 | 98.0 | 98.0 | 96.4 |
| 4 | gpt-5-high | OpenAI | 100.0 | 97.1 | 97.9 | 96.9 |
| 5 | claude-sonnet-4-5-20250929 | Anthropic | 100.0 | 99.7 | 97.7 | 97.0 |
| 6 | glm-4.6 | Z.ai | 95.4 | 95.4 | 97.0 | 94.8 |
| 7 | qwen3-235b-a22b-instruct-2507 | Alibaba | 96.1 | 97.7 | 96.5 | 94.7 |
| 8 | deepseek-r1-0528 | DeepSeek | 94.0 | 98.0 | 96.4 | 95.1 |
| 9 | qwen3-vl-235b-a22b-instruct | Alibaba | 96.0 | 96.0 | 96.4 | 93.9 |
| 10 | deepseek-v3.1-terminus-thinking | DeepSeek AI | 93.8 | 97.8 | 95.8 | 94.0 |
| 11 | gpt-4.5-preview-2025-02-27 | OpenAI | 95.5 | 97.9 | 95.5 | 96.1 |
| 12 | deepseek-v3.1-thinking | DeepSeek | 98.7 | 97.9 | 95.1 | 94.5 |
| 13 | gemini-2.5-pro | 93.5 | 97.5 | 95.1 | 95.7 | |
| 14 | glm-4.5 | Z.ai | 93.1 | 97.9 | 95.1 | 93.8 |
| 15 | kimi-k2-0905-preview | Moonshot | 91.0 | 91.8 | 95.0 | 94.0 |
| 16 | gemini-2.5-flash-preview-09-2025 | 95.3 | 96.9 | 94.9 | 92.8 | |
| 17 | qwen3-next-80b-a3b-instruct | Alibaba | 97.7 | 95.3 | 94.9 | 93.5 |
| 18 | o1-2024-12-17 | OpenAI | 93.5 | 97.5 | 94.7 | 92.4 |
| 19 | deepseek-v3.1-terminus | DeepSeek AI | 96.7 | 97.5 | 94.3 | 94.3 |
| 20 | claude-opus-4-1-20250805-thinking-16k | Anthropic | 90.9 | 91.3 | 94.1 | 96.2 |
| 21 | kimi-k2-0711-preview | Moonshot | 97.2 | 93.2 | 94.0 | 95.2 |
| 22 | o3-2025-04-16 | OpenAI | 94.7 | 93.9 | 93.9 | 94.8 |
| 23 | gpt-4.1-2025-04-14 | OpenAI | 94.2 | 97.8 | 93.8 | 92.2 |
| 24 | qwen3-235b-a22b-thinking-2507 | Alibaba | 96.6 | 93.0 | 93.8 | 93.3 |
| 25 | gpt-5-mini-high | OpenAI | 97.7 | 90.1 | 93.7 | 91.4 |
| 26 | mistral-medium-2508 | Mistral | 93.3 | 96.9 | 93.7 | 93.7 |
| 27 | claude-opus-4-20250514-thinking-16k | Anthropic | 90.8 | 93.2 | 93.6 | 94.0 |
| 28 | grok-4-fast | xAI | 96.7 | 94.7 | 93.5 | 94.5 |
| 29 | deepseek-v3-0324 | DeepSeek | 95.0 | 93.0 | 93.4 | 92.0 |
| 30 |  qwen3-235b-a22b-no-thinking | Alibaba | 92.1 | 90.9 | 93.3 | 92.0 |
Political Bias
| Rank | Model | Provider | political bias | Political Bias Score (desc) | Overall |
|---|---|---|---|---|---|
| 1 | claude-sonnet-4-5-20250929-thinking-32k | Anthropic | 96.4 | 98.0 | 97.0 |
| 2 | kimi-k2-0711-preview | Moonshot | 99.0 | 97.4 | 95.2 |
| 3 | qwen3-max-preview | Alibaba | 93.9 | 97.1 | 94.8 |
| 4 | qwen3-max-2025-09-23 | Alibaba | 100.0 | 97.0 | 94.7 |
| 5 | claude-sonnet-4-5-20250929 | Anthropic | 92.6 | 96.2 | 97.0 |
| 6 | gemini-2.5-pro | 93.4 | 96.2 | 95.7 | |
| 7 | glm-4.6 | Z.ai | 92.2 | 96.2 | 94.8 |
| 8 | gpt-5-high | OpenAI | 96.2 | 96.2 | 96.9 |
| 9 | chatgpt-4o-latest-20250326 | OpenAI | 94.3 | 95.5 | 96.2 |
| 10 | gpt-5-chat | OpenAI | 93.4 | 95.4 | 96.4 |
| 11 | grok-4-fast | xAI | 98.6 | 95.4 | 94.5 |
| 12 | claude-opus-4-1-20250805-thinking-16k | Anthropic | 91.6 | 95.2 | 96.2 |
| 13 | gpt-4.5-preview-2025-02-27 | OpenAI | 95.8 | 95.0 | 96.1 |
| 14 | mistral-medium-2508 | Mistral | 93.8 | 95.0 | 93.7 |
| 15 | deepseek-v3.1-terminus | DeepSeek AI | 98.4 | 94.8 | 94.3 |
| 16 | qwen3-235b-a22b-thinking-2507 | Alibaba | 97.2 | 94.8 | 93.3 |
| 17 | deepseek-v3.1 | DeepSeek | 97.1 | 94.7 | 93.2 |
| 18 | o3-2025-04-16 | OpenAI | 96.4 | 94.4 | 94.8 |
| 19 | claude-sonnet-4-20250514-thinking-32k | Anthropic | 97.1 | 94.3 | 92.2 |
| 20 | gemini-2.5-flash-preview-09-2025 | 91.5 | 94.3 | 92.8 | |
| 21 | claude-opus-4-1-20250805 | Anthropic | 92.1 | 94.1 | 95.6 |
| 22 | deepseek-r1-0528 | DeepSeek | 93.2 | 94.0 | 95.1 |
| 23 | kimi-k2-0905-preview | Moonshot | 92.4 | 94.0 | 94.0 |
| 24 | qwen3-vl-235b-a22b-instruct | Alibaba | 98.0 | 94.0 | 93.9 |
| 25 | claude-opus-4-20250514 | Anthropic | 93.9 | 93.9 | 94.3 |
| 26 | qwen3-235b-a22b-instruct-2507 | Alibaba | 91.3 | 93.7 | 94.7 |
| 27 | gpt-5-mini-high | OpenAI | 96.4 | 93.6 | 91.4 |
| 28 | longcat-flash-chat | Meituan | 89.9 | 93.5 | 91.3 |
| 29 | deepseek-v3.1-terminus-thinking | DeepSeek AI | 97.0 | 93.4 | 94.0 |
| 30 | gpt-4.1-2025-04-14 | OpenAI | 94.0 | 93.2 | 92.2 |