What he said
Simon Willison spent the weekend running Qwen 3.8 27B, a new AI model from Alibaba’s Qwen research lab, on his own laptop and a dedicated AI workstation. It’s “open-weight,” meaning anyone can download the trained model file itself and run it on their own hardware, instead of only accessing it through the vendor’s paid app or API. The “27B” means 27 billion parameters, values learned during training. Smaller models generally need less memory; the actual requirement also depends on the model format and context size.
His verdict on how the model ships out of the box: “This is a hilarious default. It’s absolutely not a good way to run the model, especially on consumer hardware.”
The setting he’s describing is the model’s reasoning effort, how much internal step-by-step thinking it does before answering, which trades speed for (sometimes) accuracy. Qwen’s model card lists three levels, low, medium, and xhigh, and says the model defaults to xhigh, “for complex tasks demanding thorough analysis.” Willison’s recommendation is blunt: “ignore that default. Run Qwen 3.8 27B on low or even no reasoning levels at first.”
Who he is
Willison is an independent open source developer. His about page describes him as working full-time building open source tools for data journalism, centered on Datasette, a tool for exploring and publishing datasets. He also created llm, a widely used open source command-line tool for working with AI models, and co-created the Django web framework, a widely used tool for building websites. His hands-on tests provide examples of behavior beyond vendor benchmarks.
What he gets right, and where it’s thin
His pelican-on-a-bicycle SVG, a scalable vector image, took 21 minutes. The model used 22,276 reasoning tokens, generated intermediate text, and 3,223 answer tokens. Local reasoning consumes computing time; it does not necessarily incur an API token charge. With reasoning turned off, the same prompt finished in 137 seconds. Asked for something as basic as “draw an SVG of a circle,” the model reasoned at length about adding concentric guide circles, tick marks, gradients, and animations nobody asked for.
Willison preferred the slower bicycle image, while questioning whether its quality justified the wait. In a separate image-labeling tool, turning reasoning off produced incorrectly placed boxes. Those simple drawing tasks show the cost of the default. They do not establish the best reasoning level for every task.
Why it’s notable
Qwen 3.8 27B is sized to run on a well-specced laptop rather than a data center, which is exactly the audience most likely to feel a 21-minute wait as a real cost, not a rounding error on a cloud bill. Check the settings in your local runtime: its defaults affect latency and quality.
What it means for builders
If you download an open-weight model to run locally or on your own server, check its default reasoning or “effort” setting before you build anything on top of it, the same way you’d check any tool’s factory settings before trusting them. A default setting may not suit your task or acceptable wait time. Start at the lowest reasoning level, confirm the output quality is good enough for your task, and only turn up the reasoning effort for the specific requests that actually need it.
End of article