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/bartowski/Step-3.7-Flash-GGUF:IQ1_Sollama · IQ1_S · 40.0 GiB · quantizer (bartowski)
unsloth/Step-3.7-Flash-GGUF:IQ1_Mllamacpp · IQ1_M · 52.9 GiB · quantizer (unsloth)
stepfun-ai/Step-3.7-Flash-GGUF:IQ3_XXSllamacpp · IQ3_XXS · 70.6 GiB · vendor (stepfun-ai)
hf.co/vultr/Step-3.7-Flash-GGUF:IQ3_Sollama · IQ3_S · 87.1 GiB · quantizer (vultr)
hf.co/vultr/Step-3.7-Flash-GGUF:IQ3_Mollama · IQ3_M · 87.8 GiB · quantizer (vultr)
stepfun-ai/Step-3.7-Flash-GGUF:Q3_K_Lllamacpp · Q3_K_L · 95.5 GiB · vendor (stepfun-ai)
mlx-community/Step-3.7-Flash-4bitmlx · 4BIT · 103.2 GiB · quantizer (mlx-community)
stepfun-ai/Step-3.7-Flash-GGUF:Q4_K_Sllamacpp · Q4_K_S · 103.8 GiB · vendor (stepfun-ai)
unsloth/Step-3.7-Flash-GGUF:Q4_K_Mllamacpp · Q4_K_M · 113.7 GiB · quantizer (unsloth)
stepfun-ai/Step-3.7-Flash-NVFP4vllm · NVFP4 · 120.4 GiB · vendor (stepfun-ai)
hf.co/vultr/Step-3.7-Flash-GGUF:Q4_K_Mollama · Q4_K_M · 121.5 GiB · quantizer (vultr)
cyankiwi/Step-3.7-Flash-AWQ-INT4vllm · W4A16 · 128.8 GiB · quantizer (cyankiwi)
mlx-community/Step-3.7-Flash-6bitmlx · 6BIT · 149.0 GiB · quantizer (mlx-community)
mlx-community/Step-3.7-Flash-8bitmlx · 8BIT · 194.9 GiB · quantizer (mlx-community)
stepfun-ai/Step-3.7-Flash-GGUF:Q8_0llamacpp · Q8_0 · 195.0 GiB · vendor (stepfun-ai)
stepfun-ai/Step-3.7-Flash-FP8vllm · FP8 · 197.9 GiB · vendor (stepfun-ai)
stepfun-ai/Step-3.7-Flash-GGUF:BF16llamacpp · BF16 · 367.0 GiB · vendor (stepfun-ai)