Can I run Qwen3.8-27B-MLX-6bit on an RTX 3090?
Probably, at 4-bit
The published weights are too big, but a 4-bit version (about 15.4 GB) would fit, with room for ≈ 90K tokens of context. The site hasn't analyzed a GGUF of Qwen3.8-27B-MLX-6bit yet.
Files and context on RTX 3090
Each file of Qwen3.8-27B-MLX-6bit (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.
| File | Size | On RTX 3090 | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| As published | 22.8 GB | Doesn't fit | — | — |
| 4-bit (estimate) | 15.4 GB | Fits | 90K | 181K |