AI Dev Impact Lab JA
← Rankings · AI-RELATED REPO

patrickbdevaney/deepseek-v4.1-flash-cuda-server

Pure-CUDA OpenAI-compatible server for DeepSeek-V4.1-Flash-1017 REAP-192 on Jetson AGX Thor: 25 tok/s decode, 740 tok/s prefill, Engram read from NVMe, deterministic greedy, every speedup id-exact

Cuda MIT GitHub ↗
★ 0
stars
60
AI relevance
50
solo dev
0
tool sigs

DETECTED Detected AI stack

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.

🧠 LLM providers (2)
DeepSeek OpenAI
🤖 LLM models (1)
DeepSeek

GitHub Topics

#cuda #deepseek #jetson-thor #llm-inference #openai-api

Language breakdown (by bytes)

C
1.4%
C++
16.5%
Cuda
74.1%
Shell
2.1%
Python
5.8%

Why this is classified AI-related

The AI relevance score checks four places for AI keywords and adds the weight of each one that matches. Full methodology

Repository name no match
Description +25 matched: openai
GitHub topics +25 matched: llm
README head +20 (derived from the score — README text is not stored)
Exclusion penalty -10

Total AI relevance score: 60 / 100

Position among AI repos in the same language

We track 3 Cuda repos, of which 1 score 40 or above on AI relevance. 0 have more stars than this one (top 33.3%), and 0 score higher (top 33.3%).

Owner

Account
patrickbdevaney
Type
User
Followers
37

AI & ML Engineer | LLMs, RAG, vector search, inference | MLOps, Kubernetes, Terraform | Scaling AI & bridging deep tech with real-world impact.

Dates

Created on GitHub 2026-10-30
Last push 2026-10-30
First seen here 2026-10-30
Last fetched 2026-10-30 16:10