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/ERNIE-4.5-21B-A3B-PT-GGUF:IQ1_Sollama · IQ1_S · 6.6 GiB · quantizer (unsloth)
hf.co/unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:IQ2_XXSollama · IQ2_XXS · 7.2 GiB · quantizer (unsloth)
hf.co/unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:IQ3_XXSollama · IQ3_XXS · 8.9 GiB · quantizer (unsloth)
lmstudio-community/ERNIE-4.5-21B-A3B-MLX-4bitmlx · 4BIT · 11.4 GiB · quantizer (lmstudio-community)
hf.co/unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:Q4_0ollama · Q4_0 · 11.6 GiB · quantizer (unsloth)
hf.co/unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:Q4_K_Mollama · Q4_K_M · 12.4 GiB · quantizer (unsloth)
lmstudio-community/ERNIE-4.5-21B-A3B-MLX-6bitmlx · 6BIT · 16.5 GiB · quantizer (lmstudio-community)
lmstudio-community/ERNIE-4.5-21B-A3B-MLX-8bitmlx · 8BIT · 21.6 GiB · quantizer (lmstudio-community)
hf.co/unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:Q8_0ollama · Q8_0 · 21.6 GiB · quantizer (unsloth)
hf.co/unsloth/ERNIE-4.5-21B-A3B-PT-GGUF:BF16ollama · BF16 · 40.7 GiB · quantizer (unsloth)