Benchmark
FaithJudge Hallucination Rate
Factuality benchmarks measure how often models get facts right. Scores are percentages. Lower is better.
- Models and systems measured
- 43
- Scores recorded
- 43
- Sources
- 1
- IQ weight
- 2.07
Score distribution
- 24.1%
- Mean
- 18.6%
- Median
- 16.3%
- Std dev
- 11.6%
- 10th percentile
- 46.2%
- 90th percentile
- 6.3%
- Min
This benchmark has a weight of 2.07 for chat models, measured over 31 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model or system | Score |
|---|---|---|
| 1 | Gemini 2.5 Flashgemini-2.5-flash, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 6.3% |
| 2 | Gemini 2.5 Progemini-2.5-pro, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 6.7% |
| 3 | Gemini 2.5 Progemini-2.5-pro-exp-03-25, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 7.6% |
| 4 | Gemini 2.0 Flashgemini-2.0-flash-001, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 10.2% |
| 5 | GPT 4.1gpt-4.1-2025-04-14, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 11.9% |
| 6 | GPT-4.5 Previewgpt-4.5-preview-2025-02-27, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 11.9% |
| 7 | GPT-3.5 Turbogpt-3.5-turbo-0125, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 14.9% |
| 8 | Grok 3grok-3, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 15.3% |
| 9 | GPT-4.1 minigpt-4.1-mini-2025-04-14, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 15.7% |
| 10 | GPT 4ogpt-4o-2024-11-20, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 15.9% |
| 11 | Claude Sonnet 3.7claude-3-7-sonnet-20250219, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 16.1% |
| 12 | Llama 3.3 70B InstructLlama-3.3-70B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 16.4% |
| 13 | Mistral Small 24B InstructMistral-Small-24B-Instruct-2501, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall, 2501· FaithJudge data | 17% |
| 14 | Phi 4phi-4, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 17% |
| 15 | Claude Sonnet 4claude-sonnet-4-20250514, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 18.6% |
| 16 | GPT-4o minigpt-4o-mini-2024-07-18, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 18.6% |
| 17 | Qwen2.5 32B InstructQwen2.5-32B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 19.2% |
| 18 | Claude Opus 4claude-opus-4-20250514, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 19.6% |
| 19 | Llama 4 Maverickllama-4-maverick, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 20.6% |
| 20 | Qwen2.5 72B InstructQwen2.5-72B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 20.7% |
| 21 | QwQ 32BQwQ-32B, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 24.7% |
| 22 | Llama 3.1 8B InstructLlama-3.1-8B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 28.4% |
| 23 | Qwen 2.5 14B InstructQwen2.5-14B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 29% |
| 24 | Phi 4 MiniPhi-4-mini-instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 38.4% |
| 25 | Qwen2.5 7B InstructQwen2.5-7B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 38.5% |
| 26 | jamba-1.6-miniAI21-Jamba-mini-1.6, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 39.1% |
| 27 | Llama 3.2 3BLlama-3.2-3B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 46.2% |
| 28 | Qwen2.5 3B InstructQwen2.5-3B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 56% |
| 29 | Qwen2.5 1.5B InstructQwen2.5-1.5B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 66.7% |
| 30 | Llama 3.2 1B InstructLlama-3.2-1B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 67.7% |
| 31 | Qwen2.5 0.5b InstructQwen2.5-0.5B-Instruct, o3-mini-high, FaithJudge RAG generation hallucination rate; human-example-guided LLM judge, overall· FaithJudge data | 76.3% |
Not ranked
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model or system | Score |
|---|---|
| claude-3-7-sonnet-thinking-20250219thinking, judge o3-mini-highthinking· FaithJudge data | 16.2% |
| claude-opus-4-thinking-20250514thinking, judge o3-mini-highthinking· FaithJudge data | 14.1% |
| claude-sonnet-4-thinking-20250514thinking, judge o3-mini-highthinking· FaithJudge data | 18.2% |
| o3-high-2025-04-16reasoning effort high, judge o3-mini-highreasoning effort high· FaithJudge data | 18.6% |
| o3-low-2025-04-16reasoning effort low, judge o3-mini-highreasoning effort low· FaithJudge data | 20.6% |
| o3-medium-2025-04-16reasoning effort medium, judge o3-mini-highreasoning effort medium· FaithJudge data | 17.8% |
| o3-mini-high-2025-01-31reasoning effort high, judge o3-mini-highreasoning effort high· FaithJudge data | 12.5% |
| o3-mini-low-2025-01-31reasoning effort low, judge o3-mini-highreasoning effort low· FaithJudge data | 11.9% |
| o3-mini-medium-2025-01-31reasoning effort medium, judge o3-mini-highreasoning effort medium· FaithJudge data | 11.6% |
| o4-mini-high-2025-04-16reasoning effort high, judge o3-mini-highreasoning effort high· FaithJudge data | 29.9% |
| o4-mini-low-2025-04-16reasoning effort low, judge o3-mini-highreasoning effort low· FaithJudge data | 28% |
| o4-mini-medium-2025-04-16reasoning effort medium, judge o3-mini-highreasoning effort medium· FaithJudge data | 25.8% |