Molmo 2 8B

Molmo 2 8B needs an estimated 6.5 GB at Q4, 11.1 GB at Q8 and 20.4 GB at FP16, based on the 8.7B parameter count. All figures are estimates, covering weights plus about 15% and 0.5 GB for a short context.

Yes. A 24 GB GPU fits at FP16 (estimated 20.4 GB), a 16 GB or 12 GB GPU fits at Q8 (estimated 11.1 GB) and an 8 GB GPU fits at Q4 (estimated 6.5 GB).

Yes, at Q4 with an estimated 6.5 GB. Assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits is 16 GB (estimate).

Yes, Molmo 2 8B has open weights, so you can self-host it for full control of your data.

Run Molmo 2 8B locally

Molmo 2 8B accepts text, image and video input. Audio and PDFs are not listed as input modalities.

On Arena Vision it ranks 135 (Arena's own rank) with a score of 1113, and on Arena Text it ranks 265 with a score of 1321.

Molmo 2 8B benchmarks

The closest alternatives, matched on model kind and release date with no comparable price, are AutoGLM Phone 9B Multilingual by Z.ai, Devstral Small 2 by Mistral AI and GLM-4.6V-Flash by Z.ai.

AutoGLM Phone 9B MultilingualMolmo 2 8B vs AutoGLM Phone 9B MultilingualDevstral Small 2Molmo 2 8B vs Devstral Small 2GLM-4.6V-FlashMolmo 2 8B vs GLM-4.6V-Flash

Molmo 2 8B has open weights, released under the Apache 2.0 license.

Yes, Molmo 2 8B uses the Apache 2.0 license, a permissive license that allows commercial use.

The weights can be downloaded from the Hugging Face repository allenai/Molmo2-8B.

Run Molmo 2 8B locally

Molmo 2 8B was made by Ai2 and released on December 11, 2025.

More from Ai2

Yes, Molmo 2 8B has open weights, so the model can be downloaded and self-hosted.