Pass IndexThe State of AISign in

MinerU2.5-2509-1.2B

📌 We strongly recommend using vllm for inference, as the vllm-async-engine can achieve a concurrent inference speed of 2.12 fps on one A100.

text + image → text · made by OpenDataLab

Sold by

Nobody in the catalogue publishes a price for this yet.

About

MinerU2.5-2509-1.2B — a text model from OpenDataLab.

It takes text and images and returns text. It was published in September 2025. The catalogue files it under chat. Its weights are published under the agpl-3.0 licence, which attaches conditions the plain open licences do not.

Every current figure

Maker
OpenDataLab
Register
model
Takes
text + image
Returns
text
Published
September 2025
Parameters
1.2 billion · read from its own weights
Licence
agpl-3.0

Known as 1 name

opendatalab/MinerU2.5-2509-1.2B