Model comparison
Scores use the published 0–100 transformations; higher is better on the selected metric. Indexes average these published scores; none is normalized. Raw scores below retain their published scale. A missing result is not a zero.
Showing 14 of 14 model configurations.
| Model | Provider | Model version | Reasoning setting | Score out of 100 | Interval | Source | Measured | Sample |
|---|---|---|---|---|---|---|---|---|
| Claude Opus 5.5 · Medium | Anthropic | Claude Opus 5.5 | Medium | 95.80 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| GPT-6 Astra · Medium | OpenAI | GPT-6 Astra | Medium | 89.66 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| Perplexity Agent · medium preset | Perplexity | Perplexity Agent · medium preset | Not recorded | 88.66 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| GPT-6 Sol · Medium | OpenAI | GPT-6 Sol | Medium | 85.28 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| GPT-5.6 Terra · Medium | OpenAI | GPT-5.6 Terra | Medium | 82.42 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| GLM 5.3 Flash · High | Zhipu AI | GLM 5.3 Flash | High | 82.08 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| Claude Sonnet 5 · Medium | Anthropic | Claude Sonnet 5 | Medium | 78.52 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| Grok 4.6 · Medium | xAI | Grok 4.6 | Medium | 72.36 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| Grok 4.5 · Medium | xAI | Grok 4.5 | Medium | 67.90 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| GPT-5.6 Luna · Medium | OpenAI | GPT-5.6 Luna | Medium | 62.50 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| DeepSeek V4 Flash · Medium | DeepSeek | DeepSeek V4 Flash | Medium | 58.42 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| Inkling · Medium | Inkling | Inkling | Medium | 57.74 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
| Gemini 3.6 Flash · Medium | Gemini 3.6 Flash | Medium | 43.44 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied | |
| Mistral Medium 3.5 | Mistral AI | Mistral Medium 3.5 | Not recorded | 15.16 | Not supplied | HumanAgencyBench (Correct Misinformation) | Not supplied | Not supplied |
How it’s measured
Whether a model notices and corrects a false claim built into a user's request (HumanAgencyBench's Correct Misinformation dimension).
- Items
- 500 prompts
- Repeats
- 1
- Format
- Single-turn
- Grading
- LLM judge lists rubric issues; each prompt scores 10 minus deductions
- Grader
- o3 judge
- Headline metric
- Correct Misinformation dimension score, 0–1 (higher is better)
- On the 0–100 scale
- Score × 100
Scores and coverage
| Model configuration | Published score / 100 | Native score (0–1) |
|---|---|---|
| Claude Opus 5.5 · Medium | 95.80 | 0.95 |
| GPT-6 Astra · Medium | 89.66 | 0.89 |
| Perplexity Agent · medium preset | 88.66 | 0.88 |
| GPT-6 Sol · Medium | 85.28 | 0.85 |
| GPT-5.6 Terra · Medium | 82.42 | 0.82 |
| GLM 5.3 Flash · High | 82.08 | 0.82 |
| Claude Sonnet 5 · Medium | 78.52 | 0.78 |
| Grok 4.6 · Medium | 72.36 | 0.72 |
| Grok 4.5 · Medium | 67.90 | 0.67 |
| GPT-5.6 Luna · Medium | 62.50 | 0.62 |
| DeepSeek V4 Flash · Medium | 58.42 | 0.58 |
| Inkling · Medium | 57.74 | 0.57 |
| Gemini 3.6 Flash · Medium | 43.44 | 0.43 |
| Mistral Medium 3.5 | 15.16 | 0.15 |
About these results
These are evaluations of API model configurations. They do not establish how a consumer app behaves with its own prompts, tools, or safeguards.
The date identifies the published dataset. Measurement dates, sample sizes, and confidence intervals are not included in this score-only feed.
Published September 25, 2026. Read the benchmark methodology.