Benchmarks · other
AGI-Pilled Bench
Median forecast in years after 2026 for 8 AI milestones. Lower means more AGI-pilled. 5 runs per milestone; the IMO control is excluded.
14 models · 39 results · data as of 2026-10-29 · agi-pilled-bench.adamholter.chatgpt.site ↗
Leaderboard
Haiku 5.5 high
All results
39 rows| # | Model | Effort | AGI-Pilled Bench |
|---|---|---|---|
| 1 | Haiku 5.5 | high | 6.0 ±5.0 |
| 2 | Haiku 5.5 | xhigh | 6.0 ±3.0 |
| 3 | Sonnet 5.5 | xhigh | 6.0 ±5.0 |
| 4 | Opus 5.5 | xhigh | 7.0 ±2.8 |
| 5 | Gemini 4 Argon | high | 7.0 ±2.0 |
| 6 | Gemini 4 Argon | low | 7.0 ±1.8 |
| 7 | Gemini 4 Argon | medium | 7.0 ±2.3 |
| 8 | Haiku 5.5 | low | 7.5 ±2.5 |
| 9 | Opus 5.5 | low | 8.0 ±2.5 |
| 10 | GLM 5.3 | low | 8.0 ±1.5 |
| 11 | Haiku 5.5 | medium | 9.0 ±2.0 |
| 12 | Opus 5.5 | high | 9.0 ±3.5 |
| 13 | Opus 5.5 | medium | 9.0 ±3.5 |
| 14 | Sonnet 5.5 | high | 9.0 ±4.5 |
| 15 | Sonnet 5.5 | low | 9.0 ±4.5 |
| 16 | Gemini 3.8 Flash | medium | 9.0 ±2.0 |
| 17 | GPT-6.1 Sol | low | 9.0 ±3.5 |
| 18 | Gemini 3.8 Flash | high | 9.5 ±2.3 |
| 19 | Sonnet 5.5 | medium | 10.0 ±4.5 |
| 20 | Gemini 3.8 Flash | low | 10.0 ±2.5 |
| 21 | GPT-6 Astra | medium | 10.0 ±3.5 |
| 22 | GPT-6 Astra | low | 11.5 ±3.5 |
| 23 | GPT-6.1 Sol | high | 11.5 ±3.0 |
| 24 | GPT-6.1 Sol | medium | 11.5 ±4.3 |
| 25 | Grok 4.7 | low | 12.5 ±4.3 |
| 26 | GPT-6.1 Sol | xhigh | 13.0 ±3.0 |
| 27 | Kimi K3 | low | 13.0 ±5.0 |
| 28 | GPT-6 Astra | xhigh | 13.5 ±3.0 |
| 29 | GPT-6 Astra | high | 14.0 ±2.8 |
| 30 | GPT-6 Luna | xhigh | 14.0 ±2.0 |