No machine on this network has measured this model yet. Everything below is what the catalog knows exists — where the weights come from, how large each build is, and whose work it is. Nothing here is a claim about what the model can do; that only appears once a machine has run it and signed the result.
hf.co/unsloth/Qwen3.8-27B-GGUF:IQ1_Sollama · IQ1_S · 5.8 GiB · quantizer (unsloth)
hf.co/unsloth/Qwen3.8-27B-GGUF:IQ2_XXSollama · IQ2_XXS · 6.8 GiB · quantizer (unsloth)
hf.co/unsloth/Qwen3.8-27B-GGUF:IQ3_XXSollama · IQ3_XXS · 10.2 GiB · quantizer (unsloth)
hf.co/unsloth/Qwen3.8-27B-GGUF:Q4_K_Sollama · Q4_K_S · 14.3 GiB · quantizer (unsloth)
lmstudio-community/Qwen3.8-27B-MLX-4bitmlx · 4BIT · 15.0 GiB · quantizer (lmstudio-community)
hf.co/unsloth/Qwen3.8-27B-GGUF:Q4_K_Mollama · Q4_K_M · 15.3 GiB · quantizer (unsloth)
cyankiwi/Qwen3.8-27B-AWQ-INT4vllm · W4A16 · 19.6 GiB · quantizer (cyankiwi)
RadixArk/Qwen3.8-27B-NVFP4vllm · NVFP4 · 20.4 GiB · quantizer (RadixArk)
lmstudio-community/Qwen3.8-27B-MLX-6bitmlx · 6BIT · 21.2 GiB · quantizer (lmstudio-community)
hf.co/unsloth/Qwen3.8-27B-GGUF:Q8_0ollama · Q8_0 · 27.1 GiB · quantizer (unsloth)
lmstudio-community/Qwen3.8-27B-MLX-8bitmlx · 8BIT · 27.5 GiB · quantizer (lmstudio-community)
Qwen/Qwen3.8-27B-FP8vllm · FP8 · 28.7 GiB · vendor (Qwen)
hf.co/ggml-org/Qwen3.8-27B-GGUF:BF16ollama · BF16 · 50.1 GiB · quantizer (ggml-org)
unsloth/Qwen3.8-27B-GGUF:BF16llamacpp · BF16 · 50.9 GiB · quantizer (unsloth)