ByteDance
ByteDance
Frontier

Doubao-Pro-128k

Frontier

128K-context Doubao Pro. Mid-tier of the Pro line, same model behind the 256K endpoint, sized to a more typical RAG window.

Doubao-Pro-128k is a frontier AI model from ByteDance. It costs $0.560 per million input tokens and $1.400 per million output tokens (blended $0.812/M), with a 128,000-token context window.

INPUT
$0.560/M
per million input tokens
OUTPUT
$1.400/M
per million output tokens
BLENDED 70/30
$0.812/M
unchanged since 3 May
CONTEXT
128,000
tokens
What it is good at
  • 128K context
  • Cheap Chinese chat
  • Volcengine integration
Typical use cases
  • Chinese RAG pipelines
  • Long-document chat

Benchmarks

vs. best public score
MMLU80%
Multitask academic knowledge across 57 subjects.
MATH70%
High-school competition math problems.
Python function synthesis from docstrings.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

More from ByteDance

See all 5 →

Frequently asked questions

How much does Doubao-Pro-128k cost?

Doubao-Pro-128k costs $0.560 per million input tokens and $1.400 per million output tokens, for a blended reference rate of $0.812 per million tokens.

What is Doubao-Pro-128k's context window?

Doubao-Pro-128k supports up to 128,000 tokens of context in a single request.

What is Doubao-Pro-128k best for?

Doubao-Pro-128k is well suited to 128K context, Cheap Chinese chat and Volcengine integration.

Who makes Doubao-Pro-128k?

Doubao-Pro-128k is developed and served by ByteDance.

Terms used on this page