modelbenchmark.io

Seed 2.1 Lite

doubao-seed-2-1-lite

Compare

Omnimodal ByteDance Seed 2.1 model with native text, vision, and audio understanding, 1M context, and improved coding and agent task completion

Specification

most-agreed values

Context
1M
Max output
256K
Released
2026-09-15
Knowledge cutoff
—
Retires
—
Open weights
no
Input
text, image, audio, video
Output
text

Price

US dollars per million tokens · most-agreed

Input
free
Output
free
Cache read
free

Available from 1 host

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
volcengine-coding-planfreefreefree—1M256K—