Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP requirements — can you run it?
· 9 quantizations measured · updated 2026-09-16 · source: DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF
Every quantization: real file size vs what you actually need
| Quant | File (measured) | VRAM/RAM needed (8K ctx) | Runs on (first fit) |
|---|---|---|---|
| IQ2_M | 23.8 GB | 30.1 GB | RTX PRO 6000 96GB |
| IQ3_M | 28.6 GB | 35.8 GB | RTX PRO 6000 96GB |
| Q4_K_S | 34.6 GB | 43.0 GB | RTX PRO 6000 96GB |
| Q4_K_M | 36.5 GB | 45.4 GB | RTX PRO 6000 96GB |
| Q5_K_S | 40.8 GB | 50.5 GB | RTX PRO 6000 96GB |
| Q5_K_M | 41.9 GB | 51.8 GB | RTX PRO 6000 96GB |
| IQ4_XS | 48.9 GB | 60.2 GB | RTX PRO 6000 96GB |
| Q8_0 | 60.0 GB | 73.5 GB | RTX PRO 6000 96GB |
| Q6_K | 70.0 GB | 85.6 GB | RTX PRO 6000 96GB |
KV cache uses a flat +20% context/overhead rule because this repo does not publish its config; exact numbers may differ for long contexts. MoE models keep all experts in memory — total size counts, not just active params.
Can I run Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP on my GPU? (IQ2_M recommended quant)
| Hardware | Memory | Verdict |
|---|---|---|
| RTX 3060 12GB | 12 GB | no |
| RTX 4060 Ti 16GB | 16 GB | no |
| RTX 3090 24GB | 24 GB | no |
| RTX 4090 24GB | 24 GB | no |
| RTX 5090 32GB | 32 GB | tight |
| RTX PRO 6000 96GB | 96 GB | RUNS |
| Mac 16GB unified | 16 GB | no |
| Mac 32GB unified | 32 GB | tight |
| Mac 64GB unified | 64 GB | RUNS |
| Mac 128GB unified | 128 GB | RUNS |
| Mac 256GB unified | 256 GB | RUNS |
| Mac 512GB unified | 512 GB | RUNS |
| 32GB system RAM (CPU) | 32 GB | tight |
| 64GB system RAM (CPU) | 64 GB | RUNS |
| 128GB system RAM (CPU) | 128 GB | RUNS |
| 256GB system RAM (CPU) | 256 GB | RUNS |
| 1TB server RAM (CPU) | 1024 GB | RUNS |
| 2TB server RAM (CPU) | 2048 GB | RUNS |
How much VRAM does Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP need?
The honest answer: 30.1 GB at the IQ2_M quant with 8K context — that is the measured file size (23.8 GB) plus KV cache and runtime overhead. The smallest published build needs 30.1 GB; the lossless one needs 85.6 GB.
Can I run it with CPU offload?
Yes, if combined RAM+VRAM ≥ file size — but generation speed is limited by memory bandwidth. Expect roughly: DDR5 dual channel ~10-30 tok/s for small MoE actives, single digits for big ones, 0.1-1 tok/s when experts page from disk. CPU-only is fine for batch jobs, painful for chat.
Why do other sites show different numbers?
Most pages compute weights from a formula (params × bits/8). We use the actual file sizes published in GGUF repos, which include the embedding table, unquantized tensors, and container overhead — that is why our numbers can differ from a naive calculation.
FAQ
How much VRAM does Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP need?
The measured answer: 30.1 GB at the IQ2_M quant with 8K context. The smallest published build needs 30.1 GB; the lossless (F16/BF16) build needs 85.6 GB. These are real GGUF file sizes from Hugging Face, not formula estimates.
Can I run Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP on a 24GB GPU (RTX 3090/4090)?
Not at IQ2_M (30.1 GB). Nothing fits 24GB — the smallest build needs 30.1 GB. Use the hosted API or a smaller model.
Can I run Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP on CPU with system RAM?
Yes, if your usable RAM is at least the file size. At IQ2_M you need ~30.1 GB of RAM. CPU generation is memory-bandwidth-bound: expect single-digit tok/s for large MoE models, 10-30 tok/s for small active params, and below 1 tok/s when experts page from disk.
Why do VRAM numbers for Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP differ between sites?
Most sites compute weights from a formula (params x bits/8). ModelFit uses the actual GGUF file sizes published on Hugging Face, which include the embedding table, unquantized tensors, and container overhead — so our numbers reflect what you really download and load.