Best Models
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) is currently the highest-ranked model for Humanity's Last Exam in the available data.
Use the top-ranked model as a starting point, not an automatic decision.
Compare the top choices on price, speed, availability, and your own constraints.
Open the linked result page and methodology before quoting the ranking.
Humanity's Last Exam is a broad, difficult knowledge and reasoning benchmark intended to stress advanced models.
| Rank | Model | Provider | Value |
|---|---|---|---|
| #1 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | Anthropic | 53.3% |
| #2 | GPT-5.6 Sol (max) | OpenAI | 47.2% |
| #3 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) | Anthropic | 45.7% |
| #4 | Muse Spark 1.1 (xhigh) | Meta | 45.1% |
| #5 | GPT-5.6 Sol (xhigh) | OpenAI |
This answer uses the exact same metric and live database ranking as the linked Easy Benchmarks leaderboard.
Source: Artificial Analysis Humanity's Last Exam. Updated: Jul 16, 2026.
| 44.7% |
| #6 | Gemini 3.1 Pro Preview | 44.7% |
| #7 | GPT-5.5 (xhigh) | OpenAI | 44.3% |
| #8 | GPT-5.6 Sol (high) | OpenAI | 44.1% |
| #9 | GPT-5.5 (high) | OpenAI | 43% |
| #10 | GPT-5.6 Terra (max) | OpenAI | 41.8% |
| #11 | GPT-5.4 (xhigh) | OpenAI | 41.6% |
| #12 | Gemini 3.5 Flash (high) | 41% |
| #13 | GPT-5.5 (medium) | OpenAI | 40.6% |
| #14 | Grok 4.5 (high) | SpaceXAI | 40.3% |
| #15 | GLM-5.2 (max) | Z AI | 40.1% |
| #16 | GPT-5.6 Terra (xhigh) | OpenAI | 40% |
| #17 | GPT-5.3 Codex (xhigh) | OpenAI | 39.9% |
| #18 | Muse Spark | Meta | 39.9% |
| #19 | Gemini 3.5 Flash (medium) | 39.9% |
| #20 | GPT-5.6 Sol (medium) | OpenAI | 39.7% |
