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/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:IQ1_Sollama · IQ1_S · 9.8 GiB · quantizer (bartowski)
hf.co/bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:IQ2_XXSollama · IQ2_XXS · 11.3 GiB · quantizer (bartowski)
hf.co/bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:IQ3_XXSollama · IQ3_XXS · 18.3 GiB · quantizer (bartowski)
mlx-community/Kimi-Linear-48B-A3B-Instruct-4bitmlx · 4BIT · 25.7 GiB · quantizer (mlx-community)
hf.co/bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:Q4_0ollama · Q4_0 · 26.5 GiB · quantizer (bartowski)
Firworks/Kimi-Linear-48B-A3B-Instruct-nvfp4vllm · W4A4 · 26.8 GiB · quantizer (Firworks)
hf.co/bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:Q4_K_Mollama · Q4_K_M · 28.0 GiB · quantizer (bartowski)
cyankiwi/Kimi-Linear-48B-A3B-Instruct-AWQ-4bitvllm · W4A16 · 28.4 GiB · quantizer (cyankiwi)
mlx-community/Kimi-Linear-48B-A3B-Instruct-6bitmlx · 6BIT · 37.2 GiB · quantizer (mlx-community)
mlx-community/Kimi-Linear-48B-A3B-Instruct-8bitmlx · 8BIT · 48.6 GiB · quantizer (mlx-community)
bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:Q8_0llamacpp · Q8_0 · 48.7 GiB · quantizer (bartowski)
hf.co/DhruvalLabs/Kimi-Linear-48B-A3B-Instruct-GGUF:Q8_0ollama · Q8_0 · 48.7 GiB · quantizer (DhruvalLabs)
moonshotai/Kimi-Linear-48B-A3B-Instructvllm · BF16 · 91.5 GiB · vendor (moonshotai)
bartowski/moonshotai_Kimi-Linear-48B-A3B-Instruct-GGUF:BF16llamacpp · BF16 · 91.5 GiB · quantizer (bartowski)