Written by Jakub Rusinowski · Last updated July 12, 2026
Best all-round pick: Gemma 4 27B ⭐
The NVIDIA GeForce RTX 5090 has 32 GB of VRAM, of which about 32 GB is available to a model. The largest model it can hold is Command R (35B) (35B, 21.9 GB at Q4_K_M).
| Model | Score | Memory at Q4_K_M | Context | Licence |
|---|---|---|---|---|
| Gemma 4 27B ⭐ | 96.5 | 17.1 GB | 125K | Gemma License (commercial OK) |
| Qwen 3.7 35B-A3B | 95.2 | 21.9 GB | 256K | Apache-2.0 |
| Qwen 3 32B | 95.1 | 20.6 GB | 125K | Apache 2.0 |
| Gemma 4 31B | 94.8 | 19.5 GB | 250K | Apache-2.0 |
| Qwen 3.6 35B-A3B | 94.6 | 21.9 GB | 256K | Apache-2.0 |
| Kimi K2.6 | 94.3 | 20.1 GB | 125K | Kimi License (research) |
| Workload | Recommended on this GPU |
|---|---|
| Coding | Kimi K2.5 (99.7) Qwen 3.7 35B-A3B (98.5) Qwen 3 32B (98) |
| General assistant | Gemma 4 27B ⭐ (96.5) Qwen 3.7 35B-A3B (95.2) Qwen 3 32B (95.1) |
| Reasoning | Gemma 4 27B ⭐ (99.6) GLM-4.7 / GLM-Z1 GLM-Z1 32B (Reasoning) (99.2) Qwen 3 32B (98.3) |
| RAG | Gemma 4 31B (97.5) Qwen 3.6 27B (96.2) Qwen 3.7 35B-A3B (95.2) |
| Agents | Ternary Bonsai 27B (93.3) Poolside Laguna XS 2.1 Laguna XS 2.1 33B-A3B (90.4) GLM-4.7-Flash 30B-A3B (90.3) |
| Vision | Qwen 3.7 35B-A3B (98.1) Gemma 4 31B (97.8) Qwen 3.6 35B-A3B (97.4) |
These models are strong picks generally but exceed the 32 GB this card makes available.
| Model | Needs at Q4_K_M | Short by |
|---|---|---|
| Nemotron 70B Instruct | 43.5 GB | ~11.5 GB |
| Llama 3.3 70B Instruct | 43.1 GB | ~11.1 GB |
Gemma 4 27B ⭐ is the strongest all-round pick that fits its 32 GB.
Command R (35B) — 35B parameters, needing 21.9 GB at Q4_K_M.
About 32 GB of its 32 GB, before the desktop and runtime overhead are accounted for.