Seed-Coder
ByteDance's 8B open code family where the model curates its own training data.
About
Seed-Coder (May 2025, previously Doubao-Coder) is ByteDance Seed's family of lightweight open code LLMs at 8B — base and instruct variants with 32K context plus a reasoning variant at 64K. Its headline idea is model-centric data curation: LLMs, not hand-written rules, filter and build the pretraining data. Team-reported results put it at state-of-the-art among open models of comparable size. The exact license file needs a check on the Hugging Face collection before commercial use. A credible small self-hosted coder from a lab that usually keeps models closed.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Team-reported: state-of-the-art among open-source models of comparable size across coding tasks.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: ByteDance Seed GitHub README read this session: May 8 2025, 8B base/instruct (32K) + reasoning (64K), previously Doubao-Coder, model-centric data curation, arXiv 2506.03524.; Exact license file needs HF-card confirmation before commercial use — flagged..