AI locale: esegui modelli open source sul tuo computer.

License

Apache 2.0 on the 27B; custom Qwen licence on the 2.4T MoE

Context

262k tokens

Sizes

27B dense · 2.4T-A95B MoE

August 2026, and the most-liked open-weight release of the summer. The 27B is dense, Apache 2.0, natively multimodal, and had GGUF builds from unsloth and bartowski within days — which makes it the current default for a 24 GB card. Prism ML's Bonsai 2 build squeezes the same 27B into about 6 GB, provided you run it on their llama.cpp fork. The 2.4T MoE sibling is frontier-scale and carries a different licence.

License

MIT

Context

1M tokens (65k native, extended 16x with YaRN)

Sizes

304B-A13B MoE (284B before the DSpark decoder)

The most-downloaded open-weight model on the Hub, and the largest entry here that a single machine can still load. 256 routed experts with 6 active per token means roughly 13B parameters fire on any given token, so it answers far faster than 304B suggests. The 0731 checkpoint supersedes the June preview and carries a DSpark speculative-decoding module, which is where the extra weights over the 284B base come from. Weights ship natively in FP8, so a Q8 build costs only about 7 GB more than Q4.

License

MIT

Context

1M tokens (three linear-attention layers for every sparse one)

Sizes

321B MoE (18B active)

Late August 2026, and the first natively multimodal GLM. A new base model rather than a GLM-5.2 derivative, MIT licensed, and at 5 million downloads the most pulled GLM on the Hub. unsloth's GGUF builds run from about 98 GB (UD-IQ1_M) through 109 GB (UD-Q2_K_XL) to 200 GB (UD-Q4_K_XL), which puts it in the same 128 GB class as DeepSeek V4 Flash. Not in the picker yet: the hybrid attention needs a llama.cpp pull request that has not reached a release, and we do not print commands for a build you would have to assemble by hand.

License

MIT

Context

262k tokens

Sizes

9B dense · 35B-A3B MoE · 397B MoE

August 2026. Unlike Ornith 1.0, which shipped with no model card at all, 1.5 documents its lineage (continued pretraining and reinforcement learning on top of Qwen 3.5 and Gemma 4) and publishes benchmarks, all vendor-run. The 9B and 35B-A3B keep the Qwen 3.5 architecture, so the official GGUF files run on stock llama.cpp. Download counts are very high and like counts unusually thin for the traffic, roughly one like per 7,900 downloads on the 35B GGUF; we list the model because its licence and lineage now check out, and we still cannot explain that ratio. The 397B is a single 244 GB file at Q4_K_M.

The fastest way to get a local LLM running with one command.

Maximum control and the broadest hardware coverage in the open ecosystem.

Running DeepSeek V4 Flash at usable speed on one 128 GB machine, and deliberately little else.

Browsing, comparing and chatting with local models in a desktop GUI.