Seed-2.0-Mini
Seed-2.0-mini targets latency-sensitive, high-concurrency, and cost-sensitive scenarios, emphasizing fast response and flexible inference deployment. It delivers performance comparable to ByteDance-Seed-1.6, supports 256k context, four reasoning effort modes (minimal/low/medium/high), multimodal understanding, and is optimized for lightweight tasks where cost and speed take priority.
Input
Output
Context
262K
Max Output
131K
Parameters
—
Technical Specifications
Model TypeVision
Context Window262K
Max Output Tokens131K
Parameters—
Training Cutoff—
License—
Capabilities
Input Modalities
textimagevideo
Output Modalities
text
Features
vision
Related Tools
Data sourced from official provider APIs and documentation
Last updated: May 5, 2026
Start building with the right model.
From model selection to production, one platform, no fragmentation.