The real cost to run Qwen3.6 27B NVFP4

nvidia · 18.2B parameters · 8.2K context · commercial OK

Qwen3.6 27B NVFP4 — 18.2B params (nvidia). Auto-indexed from the Hugging Face Hub (430,676 downloads). Parameter count is exact; context window, download size, and quantizations are estimates.

What it costs to run Qwen3.6 27B NVFP4 — $0 markup

• On your own machine — $0. Runs free locally if you have about 21 GB of VRAM at Q4_K_M (point Spanvero at LM Studio, Ollama, or llama.cpp).
• On your own rented GPU — from $0.26/hr. NVIDIA RTX 3090 24GB at the direct Vast.ai price (≈ $0.23 a session), $0 markup.
• Via your own API key — $0.25/1M tokens (blended input + output). No GPU to manage (rough estimate for this size).

When does owning hardware beat the API?

At $0.25/1M tokens, the cheapest API for Qwen3.6 27B NVFP4 stays cheaper than a used RTX 3090 at any daily volume — electricity alone costs more per token than the API does.

Local is never $0 — every figure here includes electricity and hardware amortization. Speeds come from published 7-8B Q4 benchmarks (your model, quant, and context will differ); break-even volumes are rounded to 2 significant figures.

Key facts

Open the free Spanvero advisor → for the live, interactive math for your exact workload and hardware.

Related models

• Qwen3.6 35B A3B NVFP4 — 18.7B, nvidia
• Gemma 4 31B IT NVFP4 — 20.9B, nvidia
• NVIDIA Nemotron 3 Super 120B A12B NVFP4 — 67.2B, nvidia
• Qwen2.5-Coder 32B Instruct — 32B, Alibaba
• gpt-oss-20b — 21B, OpenAI

Browse: More NVIDIA models · All models · Compare

Spanvero · What's new · Prices as of 2026-07-06. We're an honest advisor — $0 markup, your own accounts, we never resell compute. © 2026 Cynosure LLC.