Currently listed through these providers:
Model details
Seed 2.0 Code Preview (260328) (ByteDance)
Seed 2.0 Code Preview (260328) is a member of the ByteDance Seed 2.0 family, distinguished by a gateway-documented positioning as a variant optimized for agentic coding, frontend development, and multilingual programming. The model carries a large context window of the cataloged API limit tokens with a maximum output of the cataloged API limit tokens, giving it headroom for multi-file code reasoning and long agentic workflows. It is a closed-weight release, so practitioners interact with it through hosted inference rather than local deployment, and its listed capability set spans streaming responses, vision input, tool use, and reasoning, making it suitable for interactive coding assistants that combine visual context with code generation and external tool calls.
Practically, the model fits teams that need a single backbone for end-to-end software engineering tasks, from interpreting screenshots and design mockups to producing code across languages and orchestrating tool-driven refactors. The pairing of a broad context with explicit reasoning and tool capability tags suggests the variant is intended for structured, multi-step work such as repository-level edits, frontend scaffolding from visual references, and polyglot code translation rather than purely conversational chat. Within the Seed 2.0 family, this preview role marks it as the coding-specialized sibling, aimed at developer-facing agentic applications where long-context retention and tool integration matter more than open-weight flexibility.
Quick Info
Powered by- Provider
- LLM Gateway
- Model key
- bytedance/seed-2-0-code-preview-260328
- Release date
- Mar 28, 2026
- Last updated
- Mar 28, 2026
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $0.50
- Output token cost
- $3.00
Limits
- Output tokens
- 131,072 tokens
- Context window
- 262,144 tokens
Transparent token rates
Compare Seed 2.0 Code Preview (260328) (ByteDance) pricing
Rates are shown per one million tokens. Combined means one million input plus one million output tokens.