Seed 2.0 Lite
Language modelCard summarizedSeed2.0 Lite provides a balanced trade-off for general-purpose applications
Source: Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity (87 pp)
Use it via API
Where to call it
- BytePlus ModelArk
seed-2-0-lite-260428
Supports
- Structured outputs
- Streaming
- Reasoning control
- Tool calling
- Batch API
- Image input
- Prompt caching
Good for
- Search and deep-research agents at low cost (BrowseComp 72.1, DeepSearchQA 67.7)
- Spreadsheet and structured artifact tasks (SpreadsheetBench Verified 82.3)
- Tool use (τ²-Bench retail 90.9, BFCL-v4 72.9)
Overview
Seed 2.0 Lite
Seed2.0 Lite provides a balanced trade-off for general-purpose applications
- Cost
- $0.25 in
- Context
- 256K
- Input / output
- $0.25 / $2.00
- Max output
- 128K
- Knows up to
- —
- Released
- 14 Feb 2026
Seed2.0 Lite is the mid-size model of the Seed2.0 family, aimed at a balanced quality–cost trade-off. In ByteDance's tables it ranks first or second among the efficient models compared on agentic benchmarks (BrowseComp 72.1, SpreadsheetBench Verified 82.3) and is close to Gemini 3 Flash on knowledge and maths, but trails it on ARC-AGI-2 and long-context retrieval. §2.3, §4.4
seed-2-0-lite-260428- Input / output per 1M
- $0.25 / $2.00
- Context window
- 256K
- Max output
- 128K
- Knows up to
- —
- Reasoning
- Deep reasoning on by default; can be disabled; reasoning_effort minimal → max
- Weights
- Closed
About this card. A model card is the maker's own account of a model: what it's for, how it was trained and tested, and where it falls short. This page condenses ByteDance Seed's Seed2.0 Model Card: Towards Intelligence Frontier for Real-World Complexity. Every claim carries its source page.