Spark-v4-Ultra: benchmark results
Provider: Other.
Aggregate Rating 1498 ± 30, rank #2390 of 10916 rated models, from 55 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Reasoning - Common Sense - Chinese (CompassBench 2407) | 48.7 | Score (%) | 85.3 |
| OpenCompass Code - LCBench - English (CompassBench 2407) | 47.6 | Score (%) | 67.6 |
| OpenCompass Reasoning - Humanities - Chinese (CompassBench 2407) | 46.7 | Score (%) | 64.7 |
| OpenCompass Code - HumanEval - Chinese (CompassBench 2407) | 72.6 | Score (%) | 63.2 |
| OpenCompass Code - LCBench - Chinese (CompassBench 2407) | 37.2 | Score (%) | 61.8 |
| OpenCompass LLM - Reasoning - Chinese (CompassBench 2407) | 43.5 | Score (%) | 58.8 |
| OpenCompass Math - Arithmetic (CompassBench 2407) | 65.7 | Score (%) | 58.8 |
| OpenCompass LLM - Math - Chinese (CompassBench 2407) | 57.3 | Score (%) | 55.9 |
| OpenCompass Language - Dialogue - Chinese (CompassBench 2407) | 48.6 | Score (%) | 52.9 |
| OpenCompass Reasoning - Social - Chinese (CompassBench 2407) | 42.2 | Score (%) | 52.9 |
| OpenCompass Knowledge - Social Science - Chinese (CompassBench 2407) | 70.4 | Score (%) | 51.5 |
| OpenCompass Code - HumanEval+ - English (CompassBench 2407) | 68.9 | Score (%) | 50 |
Interactive version: theaggregate.ai/model?slug=spark-v4-ultra · How It Works · Data refreshed daily, snapshot 2026-11-08.