Aller au contenu

FR

Réserver un appel

Local AI · Creator

antirez

17 published recipes.

Recipes

1× DGX Spark

DeepSeek-V4-Flash IQ2_XXS with DwarfStar on 1× DGX Spark

19.5tok/sEveryday

1 request, prose prompt, context 4096, no speculative decoding Source

Intelligence34with thinkingOriginal model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.

≤ 3 bit: quantization may cost quality

IQ2_XXS gate/up + Q2_K down (routed, imatrix), Q8 Attn/Shared/OutDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 20, 2026

1× DGX Spark

Qwen3.8-Flash-Next IQ2_XXS with DwarfStar on 1× DGX Spark

17.1tok/sEveryday

1 request, prose prompt, context 8192, no speculative decoding Source

Intelligence40with thinkingOriginal model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.

≤ 3 bit: quantization may cost quality
License unclear

IQ2_XXS gate/up, Q2_K down; 41.73 GiB resident, 95.37 GiB BF16-n-grams auf SSDDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 20, 2026

1× Mac

GLM-5.3 UD-IQ2_XXS with DwarfStar on 1× Mac

6.14tok/sMixed

1 request, mixed prompt set, context 1k, no speculative decoding Source

Intelligence45with thinkingOriginal model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.

≤ 3 bit: quantization may cost quality
Streams from SSD
License unclear

UD-IQ2_XXS routed, blk78 Q2K, 196.58 GiBDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

Qwen3.8-Flash-Next 176.9B IQ2_XXS with DwarfStar on 1× Mac

48.7tok/sPeak

1 request, best value reported by source, context 32k Source

Intelligence40with thinkingOriginal model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.

≤ 3 bit: quantization may cost quality
License unclear

IQ2_XXS gate/up, Q2_K down; 41.73 GiB resident, 95.37 GiB BF16-n-grams auf SSDDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

Qwen3.8-Flash-Next 176.9B Q4_K with DwarfStar on 1× Mac

48.4tok/sPeak

1 request, best value reported by source, context 32k Source

Intelligence40with thinking

License unclear

Q4_K; 69.74 GiB resident, 95.37 GiB BF16-n-grams auf SSDDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

DeepSeek-V4-Flash 284B IQ2_XXS with DwarfStar on 1× Mac

39.4tok/sPeak

1 request, best value reported by source, context 2k Source

Intelligence34with thinkingOriginal model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.

≤ 3 bit: quantization may cost quality

IQ2_XXS gate/up + Q2_K down (routed, imatrix), Q8 Attn/Shared/Out, F16/F32 RestDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

DeepSeek-V4-Flash Q2 with DwarfStar on 1× Mac (M4 Max)

26.8tok/sPeak

1 request, best value reported by source, context 2k Source

Intelligenceno independent value

≤ 3 bit: quantization may cost quality

vermutlich Flash Q2 (CSV nennt Datei nicht)DwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

GLM-5.3-Flash Q4_K with DwarfStar on 1× Mac (M3 Ultra)

24.7tok/sPeak

1 request, best value reported by source, context 2k Source

Intelligence42with thinking

License unclear

Q4_K mit Q8 KDA/headDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

DeepSeek-V4-Flash Q2 with DwarfStar on 1× Mac (M2 Ultra)

23.2tok/sPeak

1 request, best value reported by source, context 2k Source

Intelligenceno independent value

≤ 3 bit: quantization may cost quality

vermutlich Flash Q2 (CSV nennt Datei nicht)DwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

2× DGX Spark

DeepSeek-V4.1-Flash Q2 with DwarfStar on 2× DGX Spark

21.9tok/sPeak

1 request, best value reported by source, context 1024 Source

Intelligence40with thinking no thinking 25

≤ 3 bit: quantization may cost quality

Q2 (imatrix), Experten hälftig je Spark; vollständige GGUF auf beiden SSDsDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 20, 2026

1× MacDeepSeek-V4-Flash (Vision Experimental)

DeepSeek-V4-Flash MXFP4 with DwarfStar on 1× Mac

19.3tok/sPeak

1 request, best value reported by source, context 12k Source

Intelligenceno independent value

Experimental
Streams from SSD

MXFP4 routed experts (nativ), Q8/F16 Rest, 145.26 GiBDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

DeepSeek-V4.1-Flash Q4 with DwarfStar on 1× Mac

18.7tok/sPeak

1 request, best value reported by source, context 4k Source

Intelligence40with thinking no thinking 25

Experimental

Q4 (imatrix routed), 294 GiB HauptgewichteDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

2× Mac

GLM-5.2 UD-IQ2_XXS with DwarfStar on 2× Mac

16.7tok/sPeak

1 request, best value reported by source, context 4k Source

Intelligence34with thinking no thinking 22Original model's value. Heavily compressed (~2.1 bit), intelligence is likely much lower.

≤ 3 bit: quantization may cost quality
License unclear

UD-IQ2_XXS routed, 196.58 GiBDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Strix Halo

DeepSeek-V4-Flash Q2 with DwarfStar on 1× Strix Halo

16.3tok/sPeak

1 request, code prompt Source

Intelligence34with thinking

≤ 3 bit: quantization may cost quality

Flash 0731 IQ2 (Q2-Recipe)DwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

GLM-5.3-Flash Q4_K with DwarfStar on 1× Mac (M5 Max)

14.9tok/sPeak

1 request, best value reported by source, context 3k Source

Intelligence42with thinking

Streams from SSD
License unclear

Q4_K, 177.77 GiBDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× Mac

DeepSeek-V4-Pro IQ2_XXS with DwarfStar on 1× Mac

12.4tok/sPeak

1 request, best value reported by source, context 2k Source

Intelligenceno independent value

≤ 3 bit: quantization may cost quality
Experimental

IQ2_XXS/Q2_K routed + Q8, imatrix 0813DwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 25, 2026

1× DGX Spark

DeepSeek-V4.1-Flash Q2 with DwarfStar on 1× DGX Spark

no documented speed

Intelligence40with thinking no thinking 25

≤ 3 bit: quantization may cost quality
Streams from SSD

Q2 (imatrix), 341-GiB-Datei inkl. 189 GiB Disk-Engram-TabellenDwarfStar (ds4) main@0aaea5a (2026-09-20)Weights Repo Updated Sep 20, 2026

← Back to overview