devs - DEV Community

devs

Nov 9

Recall.ai WASAPI Deep Dive Lands — Dual-Track Loopback Still Sets the Real Overlay Latency Floor

#windows #wasapi #audio #stt

4 min read

devs

Nov 9

sandbox-mcp Spins Disposable Windows 11 for Agents — Dual-Track WASAPI Still Owns the Call Latency Budget

#windows #wasapi #agents #audio

4 min read

devs

Nov 9

Soniox STT v6 Adds Agent Text Input — Dual-Track WASAPI Capture Still Sets the Latency Floor

#windows #audio #stt #realtime

4 min read

devs

Nov 9

Llama-modchip Puts 284B DeepSeek on 64GB CPU — Agents Still Starve on the Windows Capture Path

#windows #audio #llm #localai

2

5 min read

devs

Nov 9

Open WebUI 0.13.1 Ships Tether — Desktop Agents Still Starve on the Windows Capture Path

#windows #ai #opensource #audio

4 min read

devs

Nov 8

Process Lasso v19 LatencyTest Surfaces the DPC Spikes That Still Starve Windows Audio Capture

#windows #audio #latency #tooling

4 min read

devs

Nov 8

NamiAgent Ships a From-Scratch Local Agent Tutorial — Voice Still Dies Before the First MCP Tool Call

#agents #windows #stt #mcp

4 min read

devs

Nov 8

GitHub's @GitHub in Slack and Teams Hits GA — Chat-to-Issue Still Dies on Local Capture and Structured Parsing

#github #agents #windows #automation

4 min read

devs

Nov 8

Cursor's Real-Time Voice Chat with Agents Lands — Desktop Capture Latency Still Sets the Real Budget

#windows #audio #stt #agents

5 min read

devs

Nov 8

Speculative Listening Cuts Full-Duplex Response Start to 74-241 ms — Idle Compute in Chunked Speech Models

#windows #audio #stt #latency

4 min read

devs

Nov 7

Apiiro’s Claude Code Scan Found 2,539 Secrets in .claude/settings.local.json — Local Desktop Tools Keep Credentials Off the Git Path

#claudecode #security #windows #localai

3 min read

devs

Nov 7

Hugging Face’s 70k-Word MoE Handbook After 800k GPU Hours — Why Sparsity Still Wins for Local Real-Time Tools

#moe #inference #localai #windows

4 min read

devs

Nov 7

Agent Substrate + gVisor: Userspace Kernel for Sub-500ms Agent Resume and 10 Density

#agents #kubernetes #sandbox #gvisor

4 min read

devs

Nov 7

Intelligent Terminal 0.3 Adds Vertical Agent Sidebar and Headless Tabs — ConPTY + ACP Latency Budget on Windows

#windows #terminal #agents #conpty

4 min read

devs

Nov 7

Qwen3-ASR 1.7B Q4_K_M GGUF at 1.32 GB — Local CPU STT Now Fits Inside the Overlay Process

#stt #windows #localai #qwen

4 min read

devs

Nov 6

Qwen3 ASR 1.7B Hits 35 ms Time-to-Final on Coval — Windows Capture Still Owns the Real-Time Budget

#stt #windows #latency #qwen

4 min read

devs

Nov 6

agent-desktop 1.2.0 Turns the Accessibility Tree into 67 MCP Tools — Overlay Latency Still Wins

#mcp #windows #automation #agents

5 min read

devs

Nov 6

Foundry Local + Qwen3.8-27B Turns GitHub Copilot Local — Windows Audio Still Owns the Real-Time Budget

#windows #localai #githubcopilot #llm

4 min read

devs

Nov 5

llama.cpp 0.9.0 Ships Chatterbox TTS and OCP FP8 — Local Speech Still Dies on the Windows Capture Path

#windows #audio #llm #localai

4 min read

devs

Nov 4

Anthropic FlashProve: zkSNARK Model Provenance and the Unchanged Windows Capture Latency Problem

#windows #audio #stt #llm

6 min read

devs

Nov 4

Soniox STT v6 Agent Text Injection: Residual Echo Filtering and Windows Capture Latency Budgets

#stt #windows #audio #realtime

5 min read

devs

Nov 3

REA Crosses 400k Stars — Agents Decompile Windows Binaries Through MCP and Ghidra

#reverseengineering #windows #mcp #ai

4 min read

devs

Nov 3

Mistral Large 4 Open Weights Drop — Your Local Overlay Still Lives or Dies on the Capture Path

#mistral #localai #windows #stt

4 min read

devs

Nov 3

Soniox STT RT v5 Hits 45 ms Time-to-Final — Capture Budget Still Dominates on Windows

#stt #windows #audio #latency

4 min read

devs

Nov 3

llm 0.37 Adds Multimodal Embeddings and a Cheaper Default — What That Means for Local Audio Pipelines

#llm #embeddings #audio #windows

4 min read

devs

Nov 3

Uber’s MCP Gateway Keeps 5,000 Tools Out of the Context Window

#mcp #agents #architecture #tooling

1

4 min read

devs

Nov 2

REA Hits 420k Stars: Local MCP for Agents to Disassemble Binaries with Evidence

#reverseengineering #mcp #windows #automation

4 min read

devs

Nov 2

NVIDIA Nemotron 3.5 Transcribe: One Offline Checkpoint for Prompted Transcription, Speakers, and Translation

#nvidia #asr #stt #windows

1

4 min read

devs

Nov 2

DeepSeek dsh-superagent: Router-Worker Isolation for Parallel Coding Sessions

#agents #opensource #architecture #windows

5 min read

devs

Nov 2

Vercel agent-desktop Ships Computer Use via UIA Trees, Not Pixels

#windows #uiautomation #aiagents #automation

4 min read

devs

Nov 2

Same Haiku 5.5 Scores 82% or 65% on MCP Servers: The Harness Decides

#ai #mcp #codingagents #realtime

5 min read

devs

Nov 1

Nemotron 3.5 Transcribe: One-GPU Promptable Offline STT for Multilingual Call Pipelines

#nvidia #stt #transcription #windows

4 min read

devs

Nov 1

Azure LLM Speech 09-2026: 25 New Languages and 2 Faster Fast Transcription for Multilingual Call Pipelines

#azure #speech #stt #windows

5 min read

devs

Nov 1

agent-desktop 1.0: Observe-Act-Verify Loops via Windows UI Automation Instead of Screenshot Guessing

#windows #automation #a11y #agents

1

4 min read

devs

Nov 1

Gemini Computer Use GA: Screenshot Loops, Coordinate Grids, and Why Accessibility Trees Still Win on Windows Latency

#ai #windows #tooling #agents

3 min read

devs

Nov 1

Factory Droid 1.0 Script Mode: Hundreds of Tool Calls from One Model Request in a JS Sandbox

#ai #agents #javascript #automation

4 min read

devs

Oct 31

Microsoft’s MAI-Transcribe-2-Streaming and the 100 ms WASAPI Capture Budget

#windows #audio #stt #realtime

5 min read

devs

Oct 31

agent-desktop: Driving Windows UIs from an Agent via Accessibility Trees Instead of Screenshots

#automation #windows #aiagents #opensource

4 min read

devs

Oct 30

Mistral Large 4 Open Weights: Running a 1T MoE Locally Without the Cloud Tax

#ai #opensource #localmodels #windows

5 min read

devs

Oct 30

Hugging Face AnySpeechRecognizer: swap Apple, Parakeet or Qwen3-ASR without rewriting your STT path

#speech #localai #swift #huggingface

4 min read

devs

Oct 30

Ollama 0.40.13 puts Liquid LFM3 audio on Windows: local STT/TTS that fits in 2 GB

#ollama #localai #speech #windows

1

5 min read

devs

Oct 30

NVIDIA Nemotron 3.5 Transcribe is offline on purpose: fitting a 4.3B-active speech LLM into a live Windows call path

#nvidia #stt #windows #realtime

6 min read

devs

Oct 30

Realtime STT at $0.36/hour is not the hard part: turn-taking and the Windows capture path

#windows #audio #stt #realtime

4 min read

devs

Oct 29

BYOK for a live call assistant: why the app never sees your provider keys

#security #windows #python #privacy

1

1 min read

devs

Oct 29

Storing users' API keys in a Windows desktop app with DPAPI, from Python, without a dependency

#python #windows #security #desktop

4 min read

devs

Oct 28

Floating overlay that stays out of screen-share captures on Windows

#windows #electron #ui #desktop

1 min read

devs

Oct 28

Building a dual-track audio pipeline on Windows: mic + system audio without joining the call

#windows #audio #rust #realtime

2 min read

devs

Oct 28

Stop hammering a rate-limited LLM API: per-provider cooldowns in a live assistant

#python #llm #api #ai

2

5

4 min read

devs

Oct 28

Your WAV file says it's empty: keeping recordings readable when the process gets killed

#rust #python #audio #debugging

1

5 min read

devs

Oct 27

Two streaming speech-to-text bugs that left no error: binary frames and keyterm budgets

#rust #ai #websockets #debugging

1

4 min read

devs

Oct 27

Product screenshots from the real UI, without real user data: Electron, a mock API and capturePage

#electron #javascript #webdev #testing

3 min read

devs

Oct 27

Serving Markdown to AI agents from a Cloudflare Worker, without a second copy of the site

#cloudflare #webdev #ai #javascript

1

4 min read

devs

Oct 26

Capturing the mic and the call as two separate tracks on Windows with cpal and WASAPI loopback

#rust #windows #audio #performance

5 min read

devs

Oct 26

Where the time goes in a live AI call assistant: my code takes 150 ms, the model takes the rest

#ai #llm #performance #python

1

4 min read

Writing Debut

Want to connect with devs?