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.
bartowski/Hy3-GGUF:IQ1_Sllamacpp · IQ1_S · 59.7 GiB · quantizer (bartowski)
bartowski/Hy3-GGUF:IQ2_XXSllamacpp · IQ2_XXS · 76.5 GiB · quantizer (bartowski)
hf.co/AngelSlim/Hy3-GGUF:IQ1_Mollama · IQ1_M · 83.3 GiB · quantizer (AngelSlim)
bartowski/Hy3-GGUF:IQ3_XXSllamacpp · IQ3_XXS · 117.4 GiB · quantizer (bartowski)
hf.co/AngelSlim/Hy3-GGUF:Q3_K_Mollama · Q3_K_M · 125.4 GiB · quantizer (AngelSlim)
hf.co/vultr/Hy3-GGUF:IQ3_Sollama · IQ3_S · 130.3 GiB · quantizer (vultr)
AngelSlim/Hy3-GPTQ-Int4vllm · GPTQ · 156.9 GiB · quantizer (AngelSlim)
bartowski/Hy3-GGUF:Q4_0llamacpp · Q4_0 · 158.6 GiB · quantizer (bartowski)
tiyuvta/Hy3-NVFP4vllm · NVFP4 · 168.4 GiB · quantizer (tiyuvta)
bartowski/Hy3-GGUF:Q4_K_Mllamacpp · Q4_K_M · 169.6 GiB · quantizer (bartowski)
hf.co/AngelSlim/Hy3-GGUF:Q4_K_Mollama · Q4_K_M · 169.8 GiB · quantizer (AngelSlim)
tencent/Hy3-FP8vllm · FP8 · 279.3 GiB · vendor (tencent)
bartowski/Hy3-GGUF:Q8_0llamacpp · Q8_0 · 295.8 GiB · quantizer (bartowski)
tencent/Hy3vllm · BF16 · 556.5 GiB · vendor (tencent)