Olmo Hybrid 7B Instruct

An estimated 5.7 GB at Q4, 9.6 GB at Q8 and 17.6 GB at FP16. All VRAM figures are estimates from the 7.4B parameter count, with weights plus about 15% and 0.5 GB for a short context.

Yes. A 24 GB GPU fits at FP16 (estimated 17.6 GB), a 16 GB or 12 GB GPU fits at Q8 (estimated 9.6 GB) and an 8 GB GPU fits at Q4 (estimated 5.7 GB).

Yes. Q4 needs an estimated 5.7 GB, and assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits is 8 GB.

Yes, Olmo Hybrid 7B Instruct has open weights, so you can self-host it for full control of your data.

Run Olmo Hybrid 7B Instruct locally

No. Its input modalities are text only.

Qwen3.5 0.8B by Alibaba, Qwen3.5 2B by Alibaba and LFM2 24B A2B by Liquid AI. They are matched on model kind and release date, with no comparable price.

Qwen3.5 0.8BOlmo Hybrid 7B Instruct vs Qwen3.5 0.8BQwen3.5 2BOlmo Hybrid 7B Instruct vs Qwen3.5 2BLFM2 24B A2BOlmo Hybrid 7B Instruct vs LFM2 24B A2B

Olmo Hybrid 7B Instruct is open weights, released under the Apache 2.0 license.

Yes. Its Apache 2.0 license is a permissive license that allows commercial use.

From the Hugging Face repository allenai/Olmo-Hybrid-Instruct-DPO-7B.

Run Olmo Hybrid 7B Instruct locally

Ai2. It was released on March 5, 2026.

More from Ai2

Yes, it is open weights.