Blip2 Opt 2.7b is a vision-language model hosted on Hugging Face under the Salesforce organization. It carries the pipeline tag image-text-to-text and belongs to the blip-2 model family. The model combines visual understanding with language generation to support tasks such as visual question answering and image-to-text conversion.
It is distributed as an open model repository that includes PyTorch and safetensors weights. The library_name field identifies it as compatible with the Transformers library, allowing integration into standard machine-learning workflows that rely on this framework. Model files are publicly accessible without gating.
The repository records more than 17 million all-time downloads and approximately 500 thousand recent downloads. It has accumulated 446 likes from the Hugging Face community. The model card was first created in February 2023 and received its most recent update in February 2025.
No pricing, licensing terms, or specific training details appear in the available metadata. The entry contains no further description of architecture components, supported input formats, or example use cases beyond the pipeline and tag information.
Blip2 Opt 2.7b is a Foundation models & chat product. It focuses on bridging vision and language understanding to enable image-based question answering and captioning. It is built as an open-source project for developers. Blip2 Opt 2.7b is open source under the BSD-3-Clause license. It runs on the web, the command line, and API.
It is developed by Salesforce (United States), and the product first shipped in 2022. Development happens publicly on GitHub with 11.3k stars and 1 commits in the last 90 days. Key capabilities include Visual Question Answering, Image Captioning, and vision-Language. It exposes integrations via a public API.
Latest indexed changes and source events
Salesforce/blip2-opt-2.7b verified by the PulseGate indexer
Other apps tracked under the same category.
