Footer

Your go-to destination for discovering trending open-source projects and uncovering the insights that matter

Legal

  • Terms and conditions
  • Privacy policy

More

  • Pricing
  • Advertise
  • onedump
  • gosnakego
  • Contact us

Social

  • X (Twitter)
  • GitHub
Trendshift owl
What they build speaks louder than words
© 2026 Trendshift, created by Julian Li with ❤️
SignalAdvertise
Trendshift
SignalAdvertise
Sign in
Trendshift owl

Discover trends that matter

SignalAdvertise
DailyWeeklyMonthlyYearly
Live mentionsTopics
RepositoriesDevelopers
InsightsStats

It's a plan
It's a plan
OfinTech Desk
OfinTech Desk
Omnigraph
Omnigraph
Bifrost
Bifrost
Busbar
Busbar
Trina AX
Trina AX

divagr18/qwenputer

A computer, compiled into the weights of a language model. An exact CPU inside Qwen3.5-4B, with Gated DeltaNet state as RAM.

#Language model
Visit GitHub
Python
844
117
1 contributors
Custom license
FeaturedAdvertise to 180K+ monthly visitors
It's a plan

Open-source Linear alternative

An issue tracker where AI agents take the work you assign them

OfinTech Desk

An open-source AI workspace for finance and accounting teams, in which every figure is traced to its source

Omnigraph

Versioned graph database on object storage. Git-style workflows and multimodal retrieval

Bifrost

Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

Busbar

Control what AI can do before it acts

Trina AX

A self-hosted team AI workspace for company files, with per-person folder permissions

Momentum

How fast divagr18/qwenputer is gaining stars, and where it lands on Trendshift and GitHub Trending

Activity and trending

Stars, forks and merged pull requests each day or month, lined up with the repository's rank on Trendshift Trending

Oct 27 – Nov 22, 2026

Stars +792 in 27 days

0100200
153

Forks +117 in 27 days

0102030
22

Merged PRs 0 in 27 days

01

Trendshift Trendingwhere the repository ranked each day it made the top 25

Not in the Trendshift top 25 in this period

Oct 27Oct 31Nov 4Nov 8Nov 12Nov 16Nov 20
The left and right arrow keys step from day to day; Home and End go to the first and the last

Stars, forks, merged pull requests and Trendshift Trending rank, by day
DayStars gainedForks gainedMerged PRs
Tue, Oct 27, 2026+0+00
Wed, Oct 28, 2026+21+70
Thu, Oct 29, 2026+153+220
Fri, Oct 30, 2026+80+80
Sat, Oct 31, 2026+76+100
Sun, Nov 1, 2026+56+70
Mon, Nov 2, 2026+100+110
Tue, Nov 3, 2026+78+80
Wed, Nov 4, 2026+36+60
Thu, Nov 5, 2026+27+40
Fri, Nov 6, 2026+20+10
Sat, Nov 7, 2026+27+30
Sun, Nov 8, 2026+27+40
Mon, Nov 9, 2026+21+60
Tue, Nov 10, 2026+16+10
Wed, Nov 11, 2026+11+20
Thu, Nov 12, 2026+4+40
Fri, Nov 13, 2026+6+30
Sat, Nov 14, 2026+1+30
Sun, Nov 15, 2026+2+10
Mon, Nov 16, 2026+9+00
Tue, Nov 17, 2026+4+00
Wed, Nov 18, 2026+3+30
Thu, Nov 19, 2026+3+20
Fri, Nov 20, 2026+6+10
Sat, Nov 21, 2026+3+00
Sun, Nov 22, 2026+2+00
Total+792+1170
  • 1#1 on the list
  • 7Rank that day
  • Not trending

Social mentions

Recent discussions about this repository across the web
  1. Nov 131 post
    • Qwenputer puts a working 32-bit RISC-V CPU inside @Alibaba_Qwen Qwen3.5-4B's weights, without training the added circuitry. Zhihu contributor 酱紫君 explains how logic gates, writable memory and…

      Read full on X, posted by @ZhihuFrontier

      @ZhihuFrontier · x.com

  2. 9 days earlier
  3. Nov 41 post
    • 换条路解决大模型算术不准:不是外挂 Python,而是把一颗 CPU 编进模型权重。 Qwenputer 在 Qwen3.5-4B 前加入 11 层编译生成、非训练得到的 Transformer,用 Gated DeltaNet 状态当 RAM,让 RISC-V 程序在前向计算里执行。 值得琢磨的是这种分工:语言能力靠学习,精确执行靠编译,不必把所有能力都押在训练上。它仍是实验项目,执行会消耗…

      Read full on X, posted by @vintcessun

      @vintcessun · x.com

  4. Nov 31 post
    • Also see qwenputer:

      Read full on X, posted by @sasank51

      @sasank51 · x.com

  5. Nov 21 post
    • How: npm install, Vite build, open the island demo and switch camera presets. The build passed and the Golden hour, Underwater and Flotsam scenes rendered. Catch: ran on integrated Intel graphics at…

      Read full on X, posted by @CleanFix_Dev

      @CleanFix_Dev · x.com

  6. Nov 11 post
    • Someone put a working computer inside an AI model's brain. Then ran a second AI on it. This is Qwenputer by @divagr1925. He compiled a full RISC-V CPU straight into the weights of Qwen3.5-4B. No…

      Read full on X, posted by @Yashpatil

      @Yashpatil · x.com

  7. Oct 313 posts
    • Qwen ambassador Divyansh Agrawal compiled a 32-bit RISC-V CPU into Qwen3.5-4B, using its Gated DeltaNet state as RAM. It runs C and MicroPython in the forward pass and computes instead of guessing.…

      Read full on X, posted by @SadraMajidi04

      @SadraMajidi04 · x.com

    • 很硬核的实验:把一颗 32 位 RISC-V CPU 编译进 Qwen3.5-4B 的权重里 前面加 11 层 transformer,权重全是算出来的,不是训练出来的。大数乘法的准确率从 55% 提到 97.5%,而且全程不调用外部工具。

      Read full on X, posted by @DropRadar_io

      @DropRadar_io · x.com

    • 很硬核的实验:把一颗 32 位 RISC-V CPU 编译进 Qwen3.5-4B 的权重里 前面加 11 层 transformer,权重全是算出来的,不是训练出来的。大数乘法的准确率从 55% 提到 97.5%,而且全程不调用外部工具。

      Read full on X, posted by @DropRadar_io

      @DropRadar_io · x.com

  8. Oct 292 posts
    • someone compiled a 32-bit RISC-V CPU into 11 extra layers and attached them to Qwen3.5-4B none of those weights are learned. a compiler computes every one, and Qwen's original weights stay as they…

      Read full on X, posted by @Shaan_GS

      @Shaan_GS · x.com

    • With a calculator tool, Qwen3.5-4B solved 40/40 large multiplications in Qwenputer's author-run DeepMind test. vs. 22/40 without tools on the same problems.

      Read full on X, posted by @sakurayukiai

      @sakurayukiai · x.com

You might also like

Repositories close to divagr18/qwenputer
FeaturedAdvertise to 180K+ monthly visitors
Trina AX

A self-hosted team AI workspace for company files, with per-person folder permissions

It's a plan

Open-source Linear alternative

An issue tracker where AI agents take the work you assign them

OfinTech Desk

An open-source AI workspace for finance and accounting teams, in which every figure is traced to its source

Omnigraph

Versioned graph database on object storage. Git-style workflows and multimodal retrieval

Busbar

Control what AI can do before it acts

Bifrost

Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

Caldalis/MiniCPM-Math-harness

Python

built around a 2.5B math model that fits on one GPU or a laptop.

66

EthanHenrickson/math-mcp

TypeScript

A Model Context Protocol (MCP) server that provides basic mathematical and statistical functions to Large Language Models (LLMs). This server enables LLMs to perform accurate numerical calculations through a simple API.

19120

deepseek-ai/DeepSeek-Math

Python

DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

3.5k619

paradigma-inc/limite-violetto

Python

A model for high-frequency mathematical intelligence.

16614

QwenLM/Qwen2.5-Math

Python

A series of math-specific large language models of our Qwen2 series.

1.1k159

ImpossibleReality/Siffra

Rust

A fast and accurate calculator app written in Rust

557

alpertron/calculators

C

Source code of calculators hosted at https://www.alpertron.com.ar

25444

Bobliuuu/qwentpu

Python

a tpu on qwen

226

you-apps/CalcYou

Kotlin

Privacy focused calculator app built with MD3

31728