Atomic releases 14 DeepSeek-V4-Flash GGUF quants, recommends AD-IQ2_M for 128GB rigs
testingcatalog · x · 2026-08-04
Atomic released 14 GGUF quantizations of DeepSeek-V4-Flash 0731 on Hugging Face, spanning lossless BF16 down to 1-bit.
The model is a 284B-parameter MoE checkpoint trained with quantization awareness; the image compares 38 GGUF variants from 7 publishers and highlights AD-IQ2M as the best fit for 128GB hardware, with token-choice agreement reported at 83.6% versus the community’s other V4 Flash GGUFs.
Related event: Atomic Releases 14 Quantized Versions of DeepSeek V4 Flash(2 posts)→
More from Infra
- Ibiden’s AI substrate pricing surge sets up a clean earnings asymmetry — tengyanAI · 2026-08-04
- Big Tech’s OpenAI and Anthropic stakes are inflating reported earnings — Kr00ney · 2026-08-04
- Menlo Ventures says AI has entered phase 2, with infrastructure as the real opportunity — mmurph · 2026-08-04
- Podcast says AI CapEx, compute crunch, and debt-financed data centers are squeezing semis — BenBajarin · 2026-08-04
- MiniMax H3 open weights run 32 minutes down to 7.2 minutes on an L40S — ashishsanu · 2026-08-04
- Fluidstack takes its AI infrastructure dinner series to Austin and keeps hiring — MxMnr · 2026-08-04