Sapiens AINo longer listedAgnes 3.0 Flash
No source we read lists this model any more, so these figures stopped updating on 22 Sep 2026. They are kept because they were true then, not because they are true now.
BenchLeader Index
–
Benchmark results
13
Blended price
$0.075/M
Output speed
317 tok/s
Last measured results
| Benchmark | Score | Published |
|---|---|---|
| AA Intelligence Index v4.3.2 | 35.5 | 11 Sep 2026 |
| AA-LCR v1.1 | 81.0% | 11 Sep 2026 |
| AA-Omniscience v1.1 | -10.6 | 11 Sep 2026 |
| GPQA Diamond (AA) | 92.4% | 11 Sep 2026 |
| Humanity's Last Exam (AA) | 38.5% | 11 Sep 2026 |
| SciCode (AA) | 51.6% | 11 Sep 2026 |
| CritPt | 15.1% | 19 Sep 2026 |
| GDPval-AA v2.1, normalised score | 52.5% | 20 Sep 2026 |
| τ³-Banking (AA) | 47.6% | 19 Sep 2026 |
| Terminal-Bench 4.0 (AA) | 7.1% | 22 Sep 2026 |
| Terminal-Bench 2.1 (AA) | 82.0% | 22 Sep 2026 |
| AA-Omniscience v1.1: accuracy | 25.5% | 22 Sep 2026 |
| AA-Omniscience v1.1: non-hallucination | 51.6% | 22 Sep 2026 |
Compared with models still measured
- vs DeepSeek V4.1 Flash 1017
- vs DeepSeek V4 Flash
- vs Gemini 3.8 Flash
- vs Step 5 Preview
- vs Muse Spark 1.3
- vs Claude Sonnet 5
- vs Gemini 3.6 Flash
- vs DeepSeek V4 Pro
- vs Claude Opus 4.6
- vs Gemini 3.5 Flash
- vs Gemini 3 Pro
- vs Claude Opus 5
- vs Claude Fable 5
- vs Grok 4.6
- vs GLM 5.3 Flash
- vs Kimi K3
- vs GPT-6 Astra
- vs Qwen3.8-Flash-Next
- vs Claude Opus 4.7
- vs Gemini 3.1 Pro
- vs GPT-5.2
- vs GLM 5.3
- vs Claude Opus 4.8
- vs Gemini 3.7 Flash
- vs Qwen3.8 Max
Released 9 Sep 2026. For models that are still measured, see the leaderboards or everything else from Sapiens AI.