Nicheloom visits: 59,847 all-time · 7,913 in the last 7 days

The opportunity tracker for new startups.

Today: 895 launches · Niche of the day: browser-based video games · See what launched today →

Janus

Go binary that runs GGUF models via Vulkan on AMD/Intel/Nvidia

Get picks like this daily. Today's launch numbers, the top launches, AI and tech news, and a daily opportunity pick, straight to your inbox.

Want to reach these readers? Sponsor an issue.

HN · launched Oct 1, 2026 · 106 upvotes (top 10%) · 0 views

Niche: llm inference and deployment tools · 191 launches, 8 in the last 7 days ▲ +60% · Explore this niche →

Category and tags

  • Horizontal
  • Model & infra
  • Developer
  • AI-native
  • Hobby / open-source project

Visit website ↗ Source listing ↗

Could you build this?

No Running GGUF LLMs via Vulkan compute shaders in pure Go requires deep systems programming, low-level graphics API engineering, and tensor math kernel optimization.

What it would actually take: A real implementation requires writing or embedding custom Vulkan compute shaders (GLSL/SPIR-V) to perform quantized matrix multiplications (GGML/GGUF formats) across different GPU architectures (AMD, Intel, Nvidia). It demands deep systems programming expertise in Go/C, GPU memory management, and low-level ML compiler internals.

Discussion

5 comments analyzed.

Competitors mentioned: vLLM, llama.cpp, sglang, ExLlama

Concerns raised: Vulkan adds high overhead on Intel hardware, Performance likely identical to upstream llama.cpp wrapper, Lack of benchmarks against competitors

Feature requests: Benchmarks against vLLM, sglang, and ExLlama

Competitors

Other products that read as similar to this one : 2334 launches clear the similarity bar, closest 8 shown.

Attention rank: #250 of 2335 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).

Launched 337 days after the earliest competitor.

The niche, labels, vibe code rating and competitor comparison on this page are analysis by Nicheloom. If you use them, please credit nicheloom.com. How you may use our data.