6 models
DeepSeek-V4-Flash 284B MoE (13B active), single-file IQ2_M merged from AtomicChat's GGUF shards so Ollama can pull it. ~33 tok/s on a 128 GB Apple Silicon Mac. Model by DeepSeek (MIT), quantization by AtomicChat.
To run DeepSeek-V4-Flash in full precision lossless, run Q4 (UD-Q4_K_XL), It is 155GB. vision tools thinking
To run DeepSeek-V4-Flash in full precision lossless, run Q3 (UD-Q3_K_M), It is 129 GB. vision tools thinking
DeepSeek-V3-Pruned-Coder-411B is a pruned version of the DeepSeek-V3 reduced from 256 experts to 160 experts, The pruned model is mainly used for code generation.
Uncensored DeepSeek-R1-0528-Qwen3-8B via Heretic: 4/100 refusals (from 89), KL 0.041. Reasoning model, MIT.
To run DeepSeek-V4-Flash in full precision lossless, run IQ1 (UD-IQ1_S), It is 82GB. vision tools thinking