Mobiinfer-Android
MobiInfer runs 35B-A3B-class MoE models fully on-device on the Qualcomm Hexagon NPU, streaming experts from flash with lookahead scheduling (LES) and DSpark speculation.
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.
Visit website ↗ Source listing ↗
More like this: all 124 launches in llm inference engines →
GITHUB · launched Oct 15, 2026 · 29 upvotes (top 51%)
- Horizontal
- Model & infra
- Developer
- AI-native
- Hobby / open-source project
Similar launches
See all 70 similar launches →Could you build this?
No This is cutting-edge systems and compiler research implementing speculative decoding and flash memory streaming for MoE models directly on Qualcomm Hexagon NPU hardware.
What it would actually take: Building this requires embedded systems and hardware-accelerated ML compiler expertise targeting Qualcomm's proprietary QNN/Hexagon SDK. It implements novel low-level systems algorithms: Lookahead Scheduling (LES) to prefetch MoE weights from flash storage to limited NPU SRAM, combined with DSpark speculative decoding algorithms written in C++/assembly. Substantial low-level knowledge of memory bandwidth hierarchies, DMA transfers, and quantization kernels is indispensable.
Competitors
Other products that read as similar to this one : 70 launches clear the similarity bar, closest 8 shown.
Attention rank: #32 of 71 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).
Launched 308 days after the earliest competitor.
- aiistream-q3.6 · github · 2026-09-29 · 11 upvotes · similarity 0.43
- Android-Eclair-3DS · github · 2026-10-01 · 38 upvotes · similarity 0.39
- nemotron-30b-cpu-256k · github · 2026-10-29 · 71 upvotes · similarity 0.39
- FastStrata · github · 2026-10-20 · 23 upvotes · similarity 0.37
- Strata · github · 2026-09-24 · 918 upvotes · similarity 0.36
- Maple-Preview · hn · 2026-08-04 · 173 upvotes · similarity 0.36
- fsr4-hexagon · github · 2026-09-17 · 11 upvotes · similarity 0.36
- turnip-sd865-z · github · 2026-10-11 · 25 upvotes · similarity 0.36
Other launches for this product
- No other launches for this product.