Seed 2.0 Pro: benchmark results
ByteDance's flagship Seed 2.0 Pro reasoning and agent model. Provider: ByteDance. Released 2026-02-16. Access: API.
Unified ELO 1674 ± 1, rank #91 of 1392 rated models, from 44 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MobileWorld - User Interaction | 61.4 | Success Rate (%) | 100 |
| MMGist | 58.2 | Macro ↑ (self-reported) | 92.3 |
| Position Bias (Lechmazur) | 28 | Order Flip % (lower is better) | 91.4 |
| MobileWorld - GUI Only | 63.2 | Success Rate (%) | 91.2 |
| MobileGym-Bench | 52 | Success Rate (%) | 88.9 |
| ZeroEval GPQA Diamond | 88.9 | GPQA Diamond Score | 86.2 |
| CC-OCR V2 | 72.15 | Average (self-reported) | 85.7 |
| Persuasion (Lechmazur) | 1.64 | Average Persuasion Strength | 85.7 |
| LLM Stats Score | 39.37 | LLM Stats Score (conservative rating) | 84.3 |
| ClawProBench | 61.07 | Final Score (self-reported) | 83.9 |
| OpenClawProBench | 68.3 | Overall Score (%) | 78.8 |
| FutureX | 42.34 | Overall Score (latest week, %) | 76.9 |
Interactive version: theaggregate.ai/model?slug=seed-2-0-pro · How It Works · Data refreshed daily, snapshot 2026-09-05.