deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe
deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe is Serve DeepSeek-V4-Flash-Vision EXL3 MixedK (full 256 experts) on one NVIDIA DGX Spark with vLLM — recipe, loader patches, and receipts - vcruz305/DeepSeek-V4-Flash-Vision-EXL3-MixedK-DGX-Spark-recipe. It is ranked #1352 on the Dev Radar, in Hosting & infra, first seen 18 days ago and shared in 3 posts (24.5k views).
GitHub - vcruz305/DeepSeek-V4-Flash-Vision-EXL3-MixedK-DGX-Spark-recipe: Serve DeepSeek-V4-Flash-Vision EXL3 MixedK (full 256 experts) on one NVIDIA DGX Spark with vLLM — recipe, loader patches, and receipts · GitHub Skip to content Navigation Menu Sign in Appearance settings Search / Sign in Sign up Appearance settings You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session. You switched accounts on another tab or window. Reload to refresh your session. Dismiss alert {{ message }} vcruz305 / DeepSeek-V4-Flash-Vision-EXL3-MixedK-DGX-Spark-recipe Public…
What people said about deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe on X
🐳 DeepSeek-V4-Flash-Vision-Exp is 305B, and I packed it into a ~95 GiB mixed-precision EXL3 release that now SERVES on one DGX Spark. Receipts + Recipe below. 🚀 K2 across the routed-expert stack. K3 only where my layer sensitivity scan ranked the extra precision highest. MTP/DSpark drafter tensors preserved from…
— @ViC305, 18 days ago · 73 likes · see the post
DeepSeek-V4-Flash-Vision EXL3 MixedK just hit 85.04 aggregate tok/s across 16 concurrent streams on ONE DGX Spark GB10. 🚀 That is 3.05× the C1 aggregate throughput, with a 117.9 tok/s burst peak, native 3-layer MTP, clean CUDA graphs and the full ~95.4 GB model on one box. ⚡🔥 This one took a few different problems…
— @ViC305, 15 days ago · 64 likes · see the post
I DID IT!! DeepSeek-V4-Flash-Vision EXL3 MixedK is now running VISION + DSpark speculative decoding together on ONE DGX Spark. 🚀🚀 19.7 tok/s text decode with DSpark3, ~2.2× my plain vision path, while image prompts still work correctly from the same vLLM process. 🔥 This is the part I really wanted to prove. Not:…
— @ViC305, 16 days ago · 67 likes · see the post
Alternatives to deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe
- boat by ASCII — ascii box is now boat ( )
- Railway — Bring context from your work and personal accounts into the same conversation. Connect your accounts in the plugin…
- Nebius Token Factory — Start building
- recipes.vllm.ai — DeepSeek's first experimental multimodal V4 model — the V4-Flash MoE backbone plus a 32-layer vision tower, 1M…
- click.alibabacloud.com — Model Studio
- serverkit — ServerKit is a lightweight, modern server control panel for managing web applications, databases, and services on your…
deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe in numbers
- Rank on the Dev Radar: #1352 of 1355
- Shared in 3 posts by 1 account: @ViC305
- 24.5k views on those posts
- First seen 18 days ago, last shared 15 days ago
- Pricing seen by Jev: open source
- Market: Hosting & infra
FAQ
What is deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe?
Serve DeepSeek-V4-Flash-Vision EXL3 MixedK (full 256 experts) on one NVIDIA DGX Spark with vLLM — recipe, loader patches, and receipts - vcruz305/DeepSeek-V4-Flash-Vision-EXL3-MixedK-DGX-Spark-recipe It was first shared on X 18 days ago and is ranked #1352 on the Dev Radar.
Is deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe free?
It is open source.
Who shared deepseek-v4-flash-vision-exl3-mixedk-dgx-spark-recipe?
1 account on X, including @ViC305, in 3 posts totalling 24.5k views.
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 12.3k posts from 4.7k X accounts over the last 21 days, 1.4k tools, 12 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-19 18:49 UTC. Full method.