The smartest model that fits in a desk-side box, and what the long-memory roles run. 176B of weights with only 6B active per token, so it moves like a small model while knowing what a large one knows — 41 tok/s on a DGX Spark. It sees text, images and video, and its cache is the cheapest in the catalogue at 24 KB per token, which is what makes long context practical on one box.
The licence, in one paragraph
Qwen Community Licence 1.0 — a small business may run it commercially. Above 100M monthly users or $20M monthly revenue you must name the model publicly, and selling model access itself, or an AI assistant product, needs a separate agreement with Qwen.
What it needs
One configuration, the one we ship and quote: weights at Q4 (4-bit), a Q8 cache at 128k context — the setup most roles actually run — plus the small runtime buffer.
Not this one: Intel Arc Pro B70 · ASRock Radeon AI PRO R9700 · AMD Ryzen AI Max+ 395 station · NVIDIA RTX 5090 · NVIDIA RTX PRO 6000 Blackwell — smaller than this model at Q4 needs.
What we change, and why
Quantisation is chosen from measured GGUF file sizes, never from nominal bits-per-weight arithmetic — k-quants run 10–17% larger than their names suggest, and a card bought on a nominal figure is a card the model does not fit.
The KV cache runs at Q8_0 by default rather than F16: half the cache for well under 1% quality cost, measured in llama.cpp (768 MiB → 408 MiB for the same cache).
Context presets start at 128k — the floor for the long-running agent work these machines are bought for — and anything past the model's native window is marked and explained, never offered as equivalent.
Prompts and tool configurations are tuned per role, including the Danish document handling the roles are bought for. What each blueprint ships is the setup we run ourselves.
Vision stays on, because the roles that run this model are the ones that read documents.
Updates we have tracked
· Multimodal
Qwen3.8-Flash-Next opens its weights, and the licence has a clause aimed at businesses like ours
125B on the label, 176 GB in practice once the separate embedding table is counted, and only 6B working per token — so it is far faster than its size suggests, and it reads text, images and video. Native context is 262k; the 1M figure needs manual configuration. Read the licence before planning on it: the Qwen Community Licence 1.0 requires a separate agreement from Qwen for anyone running a "Model as a Service or AI Work Assistant business", with no revenue or user threshold. Internal use is explicitly exempt. Our everyday models are Apache 2.0 and unaffected.