Skip to main contentThe Local Bench

Vision-language · Shanghai AI Lab

InternVL3 8B

Hugging Face: OpenGVLab/InternVL3-8B · checked 2026-10-13

Strong open VLM family from OpenGVLab.

Datasheet

Maker
Shanghai AI Lab
Parameters
7.94 BVerifiedsafetensors.total from the HF API Source
Licence
Apache-2.0Verified
Access
Public download
Task (HF pipeline)
image-text-to-text
First published on HF
2025-04-10
Last updated on HF
2025-09-11
Weights in memory
16 GB at 16-bit · 4.8 GB at ~4.85 bitsEstimateParameters × bits ÷ 8. Excludes context cache and runtime overhead; not every model has a 4-bit build.

Benchmark

Not yet measured
MMMU (val)

No comparable public score for this benchmark yet. We will not estimate one.

The description summarises the model card at https://huggingface.co/OpenGVLab/InternVL3-8B. Parameters and licence are read from the Hugging Face API; the weights-in-memory line is our arithmetic, not a measurement.