makemode
the models · eu-hosted

the models you build with.

MakeMode builds on a small set of open AI models — from the quickest to the most polished. Pick by what you need; every one runs on French infrastructure, so your work never leaves the EU whichever you choose. Here's how they differ, and how they measure up.

pick by what you need

fastest to best-looking.

The same trade-off you'll see in the model picker in the web app and demo: quicker models draft simpler pages; larger ones take longer but build more detailed, polished work. One choice, no wrong answer.

Fastest

Mistral Small 3.2

The quickest way to a first draft — great for simple pages and fast iteration.

Mistral · France
Balanced

Devstral 2

Mistral's agentic coding model — more detail, still quick. Strong on code-heavy builds.

Devstral · EU
Capable

Mistral Medium 3.5

The European flagship — detailed, well-built pages. A longer wait for higher-quality output.

Mistral · France
Best-looking

GLM 5.2

The most polished output — a strong open-weights reasoning model. Slowest to build.

open weights · EU-hosted

how to choose in opencode → — in the web app it's a picker; in the terminal it's one line of config. Same models, one key.

how they measure up

the benchmarks, honestly.

Benchmarks say what a model is good at — not where your data goes (that's identical for all of them). Here's a snapshot on coding; rankings move weekly, so we link the live boards.

coding capabilitySWE-bench Verified · reported July 2026
ModelBest forSWE-bench Verified
Mistral Small 3.2fastest · lighter draftsquick pages, simple toolsgeneral model
Devstral 2agentic coding, EUcoding-heavy builds72%
Mistral Medium 3.5EU flagshipdetailed, polished pages78%
GLM 5.2open weights, EU-hostedmost polished outputtop open-weights †
On WebDev Arena (building web front-ends), Claude leads overall; among the open, EU-hostable models, GLM and the Qwen flagship tier rank highest.
† GLM 5.2 leads open-weight models on LMArena; its makers report 62% on the harder SWE-bench Pro (a tougher test, not comparable to Verified). See the live boards: LMArena · Artificial Analysis · SWE-bench.
the same for every model

whichever you pick, it stays in France.

Sovereignty isn't a property of the model — it's how MakeMode runs all of them: on Scaleway in France, your words in, a page out, nothing leaving the EU. An open model is a file we run, not a service phoning home — so where its makers are doesn't change where your data goes. That's true for the fastest Mistral and the most polished GLM alike.

the full sovereignty story →

build with one now.

Describe a page, watch it build, publish a link — on European rails.