Ternary/1-bit compressed 27B-parameter model from PrismML, about 3.9–5.9 GB, built to run on phones and laptops.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Bonsai 27B is the largest model in PrismML's Bonsai 2 family, which spans 2B to 27B parameters. It is offered in a ternary variant (about 5.9 GB) and a 1-bit variant (about 3.9 GB). According to YouTube coverage, it is a compressed version of Qwen 3.6 27B. PrismML describes the family as bringing multimodal and agentic capabilities to local AI on personal computers, smartphones and wearables.
Creator reports say the ternary version keeps about 95% of baseline quality (80.5% vs 85% average benchmark score) and is small enough to run on a recent iPhone. One reviewer found the 1-bit model fast in a Hugging Face demo but weak at coding and unable to complete an SVG drawing test, so the 1-bit variant shows clear quality limits.
The model is aimed at on-device and efficient data-center deployment. The provided pages do not give licensing, download links or pricing details, so those remain unconfirmed.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Compressed versions of Qwen 3.6 27B: ternary at 5.9GB and 1-bit at 3.9GB, small enough for an iPhone 17 Pro. Ternary retains about 95% of baseline quality (80.5% vs 85% average benchmark score).
A 1-bit compressed 27B-parameter model about 4 GB in size, designed to run on a phone. The creator found it fast in a Hugging Face demo but unable to write good code or complete the SVG test.
Closest neighbours by meaning, not just by category.
Model · PrismML
Ternary-weight (-1, 0, +1) compression of a 27B model by PrismML, shrunk to about 5.9 GB for local use.
Model · PrismML
PrismML's 1-bit/ternary compressed Flux 2 Klein text-to-image model (~1 GB) for offline on-phone generation.
Model · Alibaba
Compact 27B-parameter model from Alibaba's Qwen family, runnable locally on a single high-end consumer GPU.
Model · Alibaba
Small 0.8B-parameter open-weights language model in Alibaba's Qwen 3.5 family, aimed at lightweight and on-device use.
Model · Alibaba
Alibaba's Qwen 3.5 large open-weights language model (~122B parameters) from the Qwen family.
Model · Alibaba
Small 0.5B-parameter open-weights base language model from Alibaba's Qwen2.5 series, Apache 2.0 licensed.
