Seed 1.6 Flash: benchmark results
ByteDance's Seed 1.6 Flash, a fast multimodal model accepting text, image and video input, with a 256K context window, optimized for high-throughput, low-latency inference. Provider: ByteDance. Released 2025-06-11. Access: API.
Aggregate Rating 1602 ± 19, rank #438 of 1635 rated models, from 50 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Multivac Evaluation | 9.45 | Avg Peer Score (0-10) | 94.4 |
| GameplayQA - Intent Identification | 63.1 | Accuracy (%) on the 148 Intent Identification questions (the | 86.7 |
| IKP - T1 | 100 | Accuracy (%) | 86.1 |
| OpenLMM Spatial - 3DSRBench (Circular) | 48.5 | CircularEval + FlipEval Accuracy (%) | 86 |
| GameplayQA - Event Recognition | 74.6 | Accuracy (%) on the 61 Event Recognition questions (the mode | 80 |
| AI for Education Pedagogy - Technology | 83.96 | Accuracy (%) | 79.3 |
| GameplayQA - Cross-Video Ordering | 48.2 | Accuracy (%) on the 117 Cross-Video Ordering questions (the | 73.3 |
| GameplayQA - Object Recognition | 72.1 | Accuracy (%) on the 70 Object Recognition questions (the mod | 73.3 |
| OpenLMM Spatial - LEGO-Puzzles (Circular) | 31.5 | CircularEval Accuracy (%) | 70.2 |
| GameplayQA - Static Object Count | 50 | Accuracy (%) on the 29 Static Object Count questions (the mo | 66.7 |
| GameplayQA - Cross-Entity Referring | 65.5 | Accuracy (%) on the 423 Cross-Entity Referring questions (th | 60 |
| GameplayQA - Sync-Referring | 61.7 | Accuracy (%) on the 207 Sync-Referring questions (the model | 60 |
Interactive version: theaggregate.ai/model?slug=seed-1-6-flash · How It Works · Data refreshed daily, snapshot 2026-11-05.