Skip to content

Doubao Seed 2.0 Pro

Released 2026-02 · reasoning · 256K tokens · 6 benchmarks

Editorial notes

Flagship Pro de Seed 2.0. Medallas de oro IMO/ICPC/CMO. LMSYS Arena #6 Text. Compite directo con GPT-5.2/Claude Opus 4.5. Scores verificados contra el Model Card oficial (Seed2.0 Model Card.pdf, Tablas 3 y 11, columna Seed2.0 Pro): MMLU-Pro 87.0, GPQA-Diamond 88.9, SWE-bench-Verified 76.5, LiveCodeBench v6 87.8. Auditoria 2026-06-08: AIME-2025 corregido 98.3->93.3 (el 98.3 venia de blogs terceros; el card oficial reporta 93.3).

Spec sheet

Company
ByteDance
Country
CN
Type
reasoning
Release
2026-02
Context
256K tokens
License
proprietary
Slug
doubao-seed-2-0-pro

Benchmarks (6)

Cite this model
BibTeX · APA

BibTeX

@misc{frontier-doubao-seed-2-0-pro,
  title  = {Doubao Seed 2.0 Pro},
  author = {{ByteDance}},
  year   = {2026},
  note   = {Frontier Benchmarks AI atlas. Accessed 2026-06-17},
  url    = {https://frontierbenchmarks.com/models/doubao-seed-2-0-pro}
}

APA

ByteDance (2026). Doubao Seed 2.0 Pro [Large language model]. Frontier Benchmarks AI. Retrieved 2026-06-17, from https://frontierbenchmarks.com/models/doubao-seed-2-0-pro

Citation reflects the atlas page, not the original model paper. For the paper, see the "Resources" section above.