Can I run Nex N2.6 on my PC?
New release: Nex N2.6 was published on Hugging Face on Nov 10, 2026 by nex-agi. Numbers update hourly as GGUF files appear. See everything new this week.
Nex N2.6 (321B, 18B active) needs about 188 GB of GPU memory at 4-bit with an 8K context. No single consumer GPU we track runs it fully in VRAM at 4-bit. With llama.cpp offload, the smallest card that can run it is the RX 7600 with 128 GB of system RAM, at an estimated 6.1–11.4 tokens/sec (slow but usable). The smallest Mac that runs it is the Mac M3 Ultra 512GB.
Nex N2.6 on every GPU and Mac
Best quant that fits fully in memory at 8K context with a speed ceiling; otherwise the llama.cpp offload plan (system RAM needed, estimated speed from DDR4-3200 to DDR5-6000). Macs and unified-memory PCs share one pool, so offload adds nothing there. How offload is estimated.
| Hardware | Memory | Verdict | Details |
|---|---|---|---|
| RX 7600 | 8 GB | Offload Q2_K, 128 GB RAM · ~6.1–11 tok/s | details |
| RTX 4060 | 8 GB | Offload Q2_K, 128 GB RAM · ~6.1–11 tok/s | details |
| RTX 5060 | 8 GB | Offload Q2_K, 128 GB RAM · ~6.4–12 tok/s | details |
| RTX 3060 Ti | 8 GB | Offload Q2_K, 128 GB RAM · ~6.4–12 tok/s | details |
| RTX 4060 Ti 8GB | 8 GB | Offload Q2_K, 128 GB RAM · ~6.1–11 tok/s | details |
| RTX 3070 | 8 GB | Offload Q2_K, 128 GB RAM · ~6.4–12 tok/s | details |
| RTX 3080 10GB | 10 GB | Offload Q2_K, 128 GB RAM · ~6.6–13 tok/s | details |
| Arc B580 | 12 GB | Offload Q2_K, 128 GB RAM · ~6.6–13 tok/s | details |
| RTX 3060 12GB | 12 GB | Offload Q2_K, 128 GB RAM · ~6.4–12 tok/s | details |
| RX 7700 XT | 12 GB | Offload Q2_K, 128 GB RAM · ~6.5–12 tok/s | details |
| RTX 5070 | 12 GB | Offload Q2_K, 128 GB RAM · ~6.7–13 tok/s | details |
| RTX 4070 | 12 GB | Offload Q2_K, 128 GB RAM · ~6.6–13 tok/s | details |
| RTX 4070 Super | 12 GB | Offload Q2_K, 128 GB RAM · ~6.6–13 tok/s | details |
| RTX 4070 Ti | 12 GB | Offload Q2_K, 128 GB RAM · ~6.6–13 tok/s | details |
| RTX 3080 12GB | 12 GB | Offload Q2_K, 128 GB RAM · ~6.8–14 tok/s | details |
| Arc A770 16GB | 16 GB | Offload Q2_K, 128 GB RAM · ~6.8–13 tok/s | details |
| RTX 5060 Ti 16GB | 16 GB | Offload Q2_K, 128 GB RAM · ~6.7–13 tok/s | details |
| RTX 4060 Ti 16GB | 16 GB | Offload Q2_K, 128 GB RAM · ~6.4–12 tok/s | details |
| RX 7800 XT | 16 GB | Offload Q2_K, 128 GB RAM · ~6.9–13 tok/s | details |
| RX 7900 GRE | 16 GB | Offload Q2_K, 128 GB RAM · ~6.8–13 tok/s | details |
| RX 9070 | 16 GB | Offload Q2_K, 128 GB RAM · ~6.9–13 tok/s | details |
| RX 9070 XT | 16 GB | Offload Q2_K, 128 GB RAM · ~6.9–13 tok/s | details |
| RTX 5070 Ti | 16 GB | Offload Q2_K, 128 GB RAM · ~7.0–14 tok/s | details |
| RTX 4070 Ti Super | 16 GB | Offload Q2_K, 128 GB RAM · ~6.9–13 tok/s | details |
| RTX 4080 Super | 16 GB | Offload Q2_K, 128 GB RAM · ~6.9–14 tok/s | details |
| RTX 5080 | 16 GB | Offload Q2_K, 128 GB RAM · ~7.0–14 tok/s | details |
| RTX 4080 | 16 GB | Offload Q2_K, 128 GB RAM · ~6.9–14 tok/s | details |
| RX 7900 XT | 20 GB | Offload Q2_K, 128 GB RAM · ~7.3–14 tok/s | details |
| RX 7900 XTX | 24 GB | Offload Q2_K, 128 GB RAM · ~7.5–15 tok/s | details |
| RTX 3090 | 24 GB | Offload Q2_K, 128 GB RAM · ~7.5–15 tok/s | details |
| RTX 4090 | 24 GB | Offload Q2_K, 128 GB RAM · ~7.5–15 tok/s | details |
| RTX 3090 Ti | 24 GB | Offload Q2_K, 128 GB RAM · ~7.5–15 tok/s | details |
| RTX 5090 | 32 GB | Offload Q3_K_M, 128 GB RAM · ~6.4–13 tok/s | details |
| DGX Spark | 128 GB | No | details |
| Ryzen AI Max+ 395 128GB | 128 GB | No | details |
| Mac M1 16GB | 16 GB | No | details |
| Mac M4 16GB | 16 GB | No | details |
| Mac M2 24GB | 24 GB | No | details |
| Mac M3 24GB | 24 GB | No | details |
| Mac M4 Pro 24GB | 24 GB | No | details |
| Mac M4 32GB | 32 GB | No | details |
| Mac M1 Pro 32GB | 32 GB | No | details |
| Mac M3 Pro 36GB | 36 GB | No | details |
| Mac M4 Pro 48GB | 48 GB | No | details |
| Mac M4 Pro 64GB | 64 GB | No | details |
| Mac M1 Max 64GB | 64 GB | No | details |
| Mac M4 Max 64GB | 64 GB | No | details |
| Mac M2 Max 96GB | 96 GB | No | details |
| Mac M3 Max 128GB | 128 GB | No | details |
| Mac M4 Max 128GB | 128 GB | No | details |
| Mac M2 Ultra 192GB | 192 GB | Q2_K only (below 4-bit) | details |
| Mac M3 Ultra 256GB | 256 GB | Q3_K_M only (below 4-bit) | details |
| Mac M3 Ultra 512GB | 512 GB | Q8_0 · ~42 tok/s | details |
Questions
How much VRAM do I need for Nex N2.6?
About 188 GB at 4-bit (Q4_K_M) with an 8K context, including KV cache and runtime overhead. Longer contexts need more; the model page has the full table by quant and context.
Why can a mixture-of-experts model run on a smaller card?
Only 18B of its 321B parameters are used per token. llama.cpp's --n-cpu-moe keeps the expert weights of some layers in system RAM while attention and the KV cache stay on the GPU, so each token reads far less from slow RAM than a dense model of the same size would.
Other popular checks: Can I run gpt-oss-120b?Can I run gpt-oss-20b?Can I run Llama 70B?Can I run Qwen3?Can I run DeepSeek R1? · Best GPU by budget · Mac vs PC
Model data: Hugging Face Hub API and GGUF file sizes. GPU specs: vendor pages. Speeds are estimates, not benchmarks. Methodology · Report an error.