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.
unsloth/MiniMax-M2.5-GGUF:IQ1_Sllamacpp · IQ1_S · 58.9 GiB · quantizer (unsloth)
unsloth/MiniMax-M2.5-GGUF:IQ2_XXSllamacpp · IQ2_XXS · 69.0 GiB · quantizer (unsloth)
hf.co/mradermacher/MiniMax-M2.5-GGUF:Q2_Kollama · Q2_K · 77.6 GiB · quantizer (mradermacher)
unsloth/MiniMax-M2.5-GGUF:IQ3_XXSllamacpp · IQ3_XXS · 86.9 GiB · quantizer (unsloth)
hf.co/mradermacher/MiniMax-M2.5-GGUF:Q3_K_Sollama · Q3_K_S · 91.9 GiB · quantizer (mradermacher)
hf.co/mradermacher/MiniMax-M2.5-GGUF:Q3_K_Lollama · Q3_K_L · 110.2 GiB · quantizer (mradermacher)
amd/MiniMax-M2.5-MXFP4vllm · QUARK · 118.6 GiB · quantizer (amd)
lmstudio-community/MiniMax-M2.5-MLX-4bitmlx · 4BIT · 119.8 GiB · quantizer (lmstudio-community)
unsloth/MiniMax-M2.5-GGUF:Q4_0llamacpp · Q4_0 · 120.6 GiB · quantizer (unsloth)
hf.co/mradermacher/MiniMax-M2.5-GGUF:Q4_K_Sollama · Q4_K_S · 121.1 GiB · quantizer (mradermacher)
cyankiwi/MiniMax-M2.5-AWQ-4bitvllm · W4A16 · 121.5 GiB · quantizer (cyankiwi)
hf.co/mradermacher/MiniMax-M2.5-GGUF:Q4_K_Mollama · Q4_K_M · 128.8 GiB · quantizer (mradermacher)
unsloth/MiniMax-M2.5-GGUF:Q4_K_Mllamacpp · Q4_K_M · 128.8 GiB · quantizer (unsloth)
nvidia/MiniMax-M2.5-NVFP4vllm · NVFP4 · 130.3 GiB · quantizer (nvidia)
lmstudio-community/MiniMax-M2.5-MLX-6bitmlx · 6BIT · 173.1 GiB · quantizer (lmstudio-community)
MiniMaxAI/MiniMax-M2.5vllm · FP8 · 214.3 GiB · vendor (MiniMaxAI)
lmstudio-community/MiniMax-M2.5-MLX-8bitmlx · 8BIT · 226.3 GiB · quantizer (lmstudio-community)
unsloth/MiniMax-M2.5-GGUF:Q8_0llamacpp · Q8_0 · 226.4 GiB · quantizer (unsloth)
unsloth/MiniMax-M2.5-GGUF:BF16llamacpp · BF16 · 426.1 GiB · quantizer (unsloth)