Models. Inhaled.
A very fast mirror of the open-weights models you were going to download anyway. Same files, same checksums, same tools — just closer.
$ export HF_ENDPOINT=https://huffingface.coThat's it. That's the integration. transformers, huggingface_hub, hf — they all just work, only faster. Anything we haven't huffed yet transparently comes from the source.
25 models
3.25 TB huffed
80 huffs served
40Â Gbps of exhale
| Model | License | Size | Huffs | Status |
|---|---|---|---|---|
| sentence-transformers/all-mpnet-base-v2 | apache-2.0 | 3.8 GB | 9 | fully huffed |
| BAAI/bge-m3 | mit | 4.6 GB | 7 | fully huffed |
| BAAI/bge-reranker-v2-m3 | apache-2.0 | 2.3 GB | 7 | fully huffed |
| zai-org/GLM-5.2 | mit | 1.51 TB | 6 | huffing… |
| openai/gpt-oss-120b | apache-2.0 | 195.8 GB | 6 | huffing… |
| openai/gpt-oss-20b | apache-2.0 | 41.3 GB | 6 | huffing… |
| sentence-transformers/all-MiniLM-L6-v2 | apache-2.0 | 976.9 MB | 6 | fully huffed |
| Qwen/Qwen3-32B | apache-2.0 | 65.5 GB | 3 | fully huffed |
| mistralai/Mistral-7B-Instruct-v0.3 | apache-2.0 | 29.0 GB | 3 | fully huffed |
| Qwen/Qwen3-8B | apache-2.0 | 16.4 GB | 3 | fully huffed |
| deepseek-ai/DeepSeek-R1-Distill-Llama-8B | mit | 16.1 GB | 3 | fully huffed |
| Qwen/Qwen2.5-7B-Instruct | apache-2.0 | 15.2 GB | 3 | fully huffed |
| Qwen/Qwen3-4B | apache-2.0 | 8.1 GB | 3 | huffing… |
| Qwen/Qwen2.5-1.5B-Instruct | apache-2.0 | 3.1 GB | 3 | fully huffed |
| Qwen/Qwen3-0.6B | apache-2.0 | 1.5 GB | 3 | fully huffed |
| Qwen/Qwen2.5-0.5B-Instruct | apache-2.0 | 999.6 MB | 3 | fully huffed |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-7B | mit | 15.2 GB | 1 | fully huffed |
| google-bert/bert-base-uncased | apache-2.0 | 3.5 GB | 1 | fully huffed |
| nomic-ai/nomic-embed-text-v1.5 | apache-2.0 | 2.2 GB | 1 | fully huffed |
| BAAI/bge-base-en-v1.5 | mit | 1.3 GB | 1 | fully huffed |
| Qwen/Qwen3-Embedding-0.6B | apache-2.0 | 1.2 GB | 1 | fully huffed |
| BAAI/bge-small-en-v1.5 | mit | 401.1 MB | 1 | fully huffed |
| zai-org/GLM-5.2-FP8 | mit | 755.7 GB | 0 | huffing… |
| nvidia/GLM-5.2-NVFP4 | mit | 464.9 GB | 0 | huffing… |
| empero-ai/Qwythos-9B-Claude-Mythos-5-1M-GGUF | apache-2.0 | 96.2 GB | 0 | huffing… |
18 of 25 fully local.
Request a model
If its license lets us redistribute it, your request primes the mirror for everyone.