Embla 1: a 10B reasoning model built to run entirely on a phone
Danish startup Sevren's 10B-A1B MoE reasoning model built to run entirely on a phone — private, offline, 4-bit weights on Hugging Face — claiming 1.77s end-to-end for a 1,024-token prompt on iPhone 18 Pro.
Why it matters
A from-scratch phone-native reasoning model trained on Denmark's Gefion supercomputer; the on-device reasoning push now has a European entrant.
What you could build with it
A developer building a private journaling app could run this model fully offline on the user's phone, so entries are summarized and searched without any personal text ever leaving the device.
Does it hold up?
Too early to judge: the latency and benchmark figures are self-measured and unverified, and no independent on-device tests have appeared yet.
Built with Embla 1
- Embla 1sevren.ai · Sevren's official Embla 1 launch post, with the phone-native design and latency claims.
- Embla 1: A Reasoning Model That Runs on Your Phone — Private, Offline, and Genuinely Fastdev.to · Independent writeup of the 57.4 benchmark average, the 1.77s iPhone latency, and the caveats — all figures are Sevren's own, unverified by any independent lab yet.
- The Architecture of Embla 1sevren.ai · Deep dive into the sparse attention, KV bridges and ReLU-squared MLPs that make the model fast on phone hardware.
Learn more
First spotted on article: source.
More AI models
Edge0-145B: open 145B MoE that runs on a single consumer GPUEdge0 open-sourced 145B and 120B latent-MoE models derived from Qwen3.8-Flash-Next, plus an inference framework that…models · JEV 0.74Strata: run a 125B model on a gaming PCFree, MIT-licensed local inference engine that runs Qwen3.8-Flash-Next (125B MoE, about 6B active per token) on…models · JEV 0.73Qwen4-Max: Alibaba's new flagship model goes liveAlibaba's 2.4T-parameter flagship is now generally available via API, with a 1M-token context, native vision, and…models · JEV 0.67smolmaxxing: run 16B LLMs on a 1 GB TV box or a 2013 iPadAn MIT-licensed llama.cpp patch set that runs language models up to a 16B MoE on devices with about 1 GB of usable RAM…models · JEV 0.65Blue TTS v3: a 388MB open text-to-speech model with voice cloningAn open-source 388MB text-to-speech model (MIT) with voice cloning from about 10 seconds of speech, 7 languages, 15…models · JEV 0.65Haru-4BOpen full-duplex spoken-dialogue foundation model that listens and speaks simultaneously (7.5 Hz frames, 133 ms per…models · JEV 0.63
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.