# Best local LLMs for Nvidia GeForce RTX 3090 (24GB)

Updated: 2026-06-15

Nvidia GeForce RTX 3090 (24GB): 24 GB vram, ~23 GB usable. Runs 58 of 67 tracked models.

- Command R 35B (35B): tight
- Yi 1.5 34B (34B): tight
- Qwen2.5 32B (32B): tight
- Qwen3 32B (32B): tight
- DeepSeek-R1-Distill-Qwen 32B (32B): tight
- Qwen2.5 Coder 32B (32B): tight
- Granite 4.0 H Small (32B): runs
- Qwen3 30B-A3B (30.5B): tight
- Sarvam-30B (30B): tight
- Gemma 2 27B (27B): runs
- Gemma 3 27B (27B): runs
- Mistral Small 3 24B (24B): runs

More: https://localmodel.run/best-llm-for/nvidia-rtx-3090-24gb