SenseNova V6 Reasoner: benchmark results
SenseTime's SenseNova V6 Reasoner (April 2025), a closed native Mixture-of-Experts multimodal foundation model with over 600 billion parameters, supporting multimodal long chain-of-thought reasoning up to 64K tokens. Provider: SenseTime. Released 2025-04-10. Access: API.
Aggregate Rating 1651 ± 37, rank #317 of 1635 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SuperCLUE General (May 2025) - Agent | 69.59 | Score | 92.3 |
| SuperCLUE General (July 2025) - Agent | 83.58 | Score | 90.2 |
| SuperCLUE General (May 2025) - Math Reasoning | 62.1 | Score | 89.7 |
| SuperCLUE General (May 2025) - Overall | 62.96 | Score | 87.2 |
| SuperCLUE General (May 2025) - Text Understanding and Creation | 80.43 | Score | 87.2 |
| SuperCLUE General (July 2025) - Code Generation | 80.2 | Score | 80.4 |
| SuperCLUE General (May 2025) - Precise Instruction Following | 31.93 | Score | 79.5 |
| SuperCLUE General (May 2025) - Science Reasoning | 47.52 | Score | 73.1 |
| LLM2014 Logic 2025-04 | 55.28 | Median Score | 71.4 |
| MathVision | 55.39 | Overall Accuracy (%) | 70.7 |
| SuperCLUE General (July 2025) - Overall | 60.73 | Score | 69.6 |
| SuperCLUE General (May 2025) - Code Generation | 86.19 | Score | 69.2 |
Interactive version: theaggregate.ai/model?slug=sensenova-v6-reasoner · How It Works · Data refreshed daily, snapshot 2026-11-05.