One stack,
end to end.
We build the full path from architecture to model. microverse searches for better building blocks; SFTSuite and taskgen turn that into curriculum training data; Claire — a 6B-A500M mixture-of-experts on a custom architecture — is the model they produce. Everything we can make open — weights on Hugging Face, code on GitHub, datasets in full — we do.
Efficient models, from scratch
Claire is our clean-sheet language model — a custom architecture and training recipe, built as a 6B mixture-of-experts with only ~500M parameters active per token. We are convinced a carefully designed sparse model punches well above its active-parameter weight: efficient enough to own and run yourself, capable enough to reason, write code and call tools.
Automated architecture discovery
microverse is our LLM-in-the-loop harness for finding better building blocks. It proposes attention mechanisms and transformer blocks, sandbox-trains each against a synthetic gauntlet, and surfaces the few structural ideas worth promoting to a real training run — before we spend the compute.
Data as curriculum
SFTSuite and taskgen treat data as a first-class part of the model. Conversation traces are generated, validated against the real tokenizer, and ordered so that position in the corpus is the curriculum — simple and short first, long and multi-turn last. The raw datasets are published openly.
Open by default
Weights, code, datasets and tools ship in the open: Abacus in your terminal, runmonitor for live training, offside-checkpoints for storage, plus our published models on Hugging Face. No waitlists. Built because we needed them; open because someone else might too.
Follow the build.
An occasional dispatch from the lab — progress on Qwythos and Claire, what we found with microverse, new Abacus releases and the one thing we got wrong that week. No hype, no roadmap theatre. Cancel from any line.