Model reference · open weights

Llama-3-Dragonfly-Med

Llama-3-Dragonfly-Med is an open-weight language model from togethercomputer. Llama-3-8B-Dragonfly-Med-v1 (BF16) weighs 16.2 GB; the smallest configuration that runs it is RTX 4090 24 GB.

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

• Llama-3-Dragonfly-Med is a multimodal biomedical visual-language model developed by Together AI, designed for image-text-to-text tasks in English.
• It is an autoregressive transformer-based model fine-tuned from Llama 3, intended for research on large visual-language models by researchers and hobbyists.
• The model operates under the Llama 3 Community License Agreement but is restricted to non-commercial use only due to the licensing of its training dataset.

Summary of the togethercomputer/Llama-3-8B-Dragonfly-Med-v1 model card, 2026-10-01

Hugging Face ↗Paper ↗