OvisOCR3
ATH-MaaS · Oct 21, 2026
OvisOCR3 is a image-text-to-text model published on Hugging Face by ATH-MaaS.
Open-weight model1,613 downloadsapache-2.0
OvisOCR2
OvisOCR2 is a image-text-to-text model published on Hugging Face by ATH-MaaS.
- Downloads (30d)
- 78,222
- Released
- Jul 13, 2026
Specifications
OvisOCR2 specs
- Provider
- ATH-MaaS
- Type
- Open-weight model
- Model ID
ATH-MaaS/OvisOCR2- Task
- image text to text
- Released
- Jul 13, 2026
- License
- apache-2.0
- Architecture
- transformers
- Base model
- Qwen/Qwen3.5-0.8B
- Downloads
- 78,222Hugging Face, last 30 days
- Likes
- 361
- Source
- Hugging Face
Overview
About OvisOCR2
OvisOCR2 is an open-weight model from ATH-MaaS, released on Jul 13, 2026, with the apache-2.0 license, 78,222 Hugging Face downloads in the last 30 days.
Specs are sourced from Hugging Face and were last checked on Aug 2, 2026. Always confirm current pricing and limits with the provider before production use.
Multimodal
FAQ
OvisOCR2 questions
When was OvisOCR2 released?
OvisOCR2 was published on Jul 13, 2026.
Can I run OvisOCR2 locally?
OvisOCR2 is published as an open-weight model on Hugging Face under the apache-2.0 license, so you can download the weights and run them on compatible hardware, subject to the license terms.