Doubao-Seed-1.6-251015 (Thinking): benchmark results
Provider: ByteDance. Access: Open.
Unified ELO 1593 ± 1, rank #443 of 2032 rated models, from 61 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Reasoning - Common - Chinese (CompassBench 2510) | 83.7 | Score (%) | 96.2 |
| OpenCompass LLM - Reasoning - English (CompassBench 2510) | 67.9 | Score (%) | 92.3 |
| OpenCompass Language - Creation - Chinese (CompassBench 2510) | 92.7 | Score (%) | 92.3 |
| OpenCompass Reasoning - Common - English (CompassBench 2510) | 77.9 | Score (%) | 92.3 |
| OpenCompass LLM - Reasoning (CompassBench 2510) | 68.5 | Score (%) | 90.4 |
| SuperCLUE General (November 2025) - Hallucination Control | 89.17 | Score | 90.3 |
| OpenCompass LLM - Reasoning - Chinese (CompassBench 2510) | 69.1 | Score (%) | 88.5 |
| OpenCompass Language - NLP - English (CompassBench 2510) | 95.1 | Score (%) | 84.6 |
| OpenCompass Math - Competition - English (CompassBench 2510) | 75 | Score (%) | 84.6 |
| OpenCompass LLM - Language - Chinese (CompassBench 2510) | 87.4 | Score (%) | 82.7 |
| OpenCompass Language - NLP - Chinese (CompassBench 2510) | 83 | Score (%) | 82.7 |
| OpenCompass Language - Instruction Following - Chinese (CompassBench 2510) | 80.8 | Score (%) | 80.8 |
Interactive version: theaggregate.ai/model?slug=doubao-seed-1-6-251015-thinking · How It Works · Data refreshed daily, snapshot 2026-09-26.