Open-weight models
30 models and 10 datasets on Hugging Face, downloaded more than 16M times. The main language models ship with GGUF builds for llama.cpp, Ollama and LM Studio on the day they come out, and PiVo runs on any CPU.
Hugging Face, November 2026. Downloads count the original weights and the GGUF builds together.
Quark
New
A small hybrid language model: three local layers under each global attention layer, trained on 500B tokens and extended to 128K context. The Raw checkpoint is the pre-anneal weights for continued pretraining; Zero is the 128K base after reinforcement learning, with no supervised fine-tuning stage. These are research artifacts, not products, and have not been safety-evaluated.
| Model | Parameters | Formats | Downloads | Released |
|---|---|---|---|---|
| Quark-150M-128K-Zero | 165M | Safetensors | New | |
| Quark-150M-128K-Base | 165M | Safetensors | New | |
| Quark-150M-32K-Base | 165M | Safetensors | New | |
| Quark-150M-32K-Raw | 165M | Safetensors | New |
PiVo
New
A complete English text-to-speech stack in 1,401,562 learned parameters: a 584,965-parameter acoustic model and an 816,597-parameter vocoder derived from FARGAN, producing 16 kHz audio in two synthetic voices. It runs on a plain CPU without PyTorch, and PiVo Studio runs it from one Rust binary. The words come through clearly; some sentences still sound metallic.
Qwen3.8 Distill
7.4M downloads
Full-parameter distillations of the Qwen3.8 frontier models into the Qwen3.5 dense and Qwen3.6 mixture-of-experts architectures. Every answer opens with a reasoning block learnt from teacher traces weighted towards mathematics and code. Native function calling and a 262,144-token context.
| Model | Parameters | Formats | Downloads | Released |
|---|---|---|---|---|
| Qwen3.8-35B-A3B-Distill | 35B total, 3B active | Safetensors, GGUF | 1,997,270 | September 2026 |
| Qwen3.8-9B-Distill | 9B | Safetensors, GGUF | 1,733,737 | August 2026 |
| Qwen3.8-4B-Distill | 4B | Safetensors, GGUF | 2,037,972 | August 2026 |
| Qwen3.8-2B-Distill | 2B | Safetensors, GGUF | 1,634,999 | August 2026 |
Qwen3.8-27B-Ridge
1.1M downloads
A quantization written for Qwen3.8's hybrid layers: the Gated-DeltaNet state path at Q8_0, the mixers at Q4_K, the bits saved in mid-stack feed-forward layers. Nothing was stripped to make the file fit; the native MTP draft head stays in the GGUF. Wiki perplexity 7.82 against 7.15 in BF16.
| Model | Parameters | Formats | Downloads | Released |
|---|---|---|---|---|
| Qwen3.8-27B-Ridge-GGUF | 27B, 11.7 GiB | GGUF | 1,133,296 | August 2026 |
Qwythos
6.8M downloads
Full-parameter reasoning models on the Qwen3.5 9B and 27B bases, trained on the same curriculum of verified chain-of-thought traces. The native multi-token-prediction head, the vision tower and a 1,048,576-token context stay intact.
| Model | Parameters | Formats | Downloads | Released |
|---|---|---|---|---|
| Qwythos-27B-v1 | 27B | Safetensors, GGUF | 885,011 | July 2026 |
| Qwythos-9B-v2 | 9B | Safetensors, GGUF | 1,949,297 | July 2026 |
| Qwythos-9B-Claude-Mythos-5-1M | 9B | Safetensors, GGUF | 3,973,274 | June 2026 |
Brewed with Brewery
Demo models
Each made start to finish in one guided session with Brewery on a single rented GPU, model card included.
| Model | Type | About | Downloads | Released |
|---|---|---|---|---|
| Homebrew-Qwen3.5-2B-Grandmas-Kitchen | 2B | Writes home-cooking recipes in a warm grandmother's voice. Made in one guided session. | 38 | October 2026 |
| Homebrew-Qwen-Image-2.1-Y2K | Image LoRA | A Qwen-Image 2.1 LoRA for the look of early-2000s digital snapshots. | 500 | October 2026 |
Earlier releases
March to June 2026
| Model | Parameters | Formats | Downloads | Released |
|---|---|---|---|---|
| Qwable-9B-Claude-Fable-5 | 9B | Safetensors, GGUF | 143,060 | June 2026 |
| Qwen3.5-9B-Claude-Code | 9B | Safetensors, GGUF | 28,046 | March 2026 |
| Qwen3.5-9B-Claude-Opus-4.6-Distill | 9B | Safetensors, GGUF | 12,109 | March 2026 |
| openNemo-9B | 9B | Safetensors | 8,809 | March 2026 |
| openNemo-9B-abliterated | 9B | Safetensors, GGUF | 15,729 | March 2026 |
| openNemo-9B-Claude-Opus-4.6-distill | 9B | Safetensors | 7,685 | March 2026 |
| openNemo-Cascade-2-30B-A3B | 30B total, 3B active | Safetensors | 7,801 | March 2026 |
Datasets
10 on Hugging Face
Training traces and task lists used for our distillations, published on Hugging Face.