AICoder

AI Coding Live News

First-hand AI coding updates from 𝕏/Twitter frontier labs and researchers, distilled by AI with entity benchmarks.

Explore live on 𝕏 ↗

AI Coding Live News

C
Cua@trycua·
👀19
🔥88
View on 𝕏 ↗

Cua + Scale AI Launch Cua Driver Bench: Background Desktop Agents vs Undisturbed Users; Vendor macOS Full-Pass ~58% for cua-driver

On 2026-11-06, Cua (@trycua) and Scale AI AgentEnv launched Cua Driver Bench (Bench v2): same Claude Sonnet 5.5, tasks, and limits with only the computer-use driver swapped, scoring task done, GUI-only, and eight undisturbed-user criteria. Scale macOS background full-pass: cua-driver ~58% (23/40), agent-desktop ~20%, browser-use ~3%; Cua-run Codex Computer Use ~33%. Tasks and trajectories are public to rerun.

Cua + Scale AI Launch Cua Driver Bench: Background Desktop Agents vs Undisturbed Users; Vendor macOS Full-Pass ~58% for cua-driver

Key Takeaways

  • Source: Cua official blog + @trycua thread on 2026-11-06; Scale ran macOS/Linux; Cua ran Windows and some controls
  • Method: 10 Bench v2 tasks; background keeps a simulated user typing; full pass = done + GUI-only + 8 undisturbed criteria
  • macOS background full-pass (Scale, vendor): cua-driver 58%, agent-desktop 20%, browser-use 3%; Codex CU (Cua-run) 33%
  • Foreground (Scale): cua-driver ~78%, agent-desktop 70%, browser-use 64%; gaps are mostly disturbance
  • Rerun: GitHub blog/libs/cua-driver-bench; HF Space/dataset earakely-scale/CuaDriverBench-AgentEnv
Read full story
Linked Models & Agents:
⚡ Claude Sonnet 5.5🤖 cua-driver🤖 codex🤖 agent-desktop🌐 Cua🌐 Scale AI🌐 OpenAI🌐 Vercel
#Cua Driver Bench#Cua#Scale AI#AgentEnv#computer use#Benchmark#Cua driver
G
Google AI StudioGoogleAIStudio·
👀21
🔥86
View on 𝕏 ↗

Gemini API Computer Use Tool Goes GA: Browser, Mobile and Desktop Agents Production-Ready, No Breaking Changes from Preview

The Gemini API changelog shows the Computer Use tool reached general availability on October 30 across both the Interactions and GenerateContent APIs; @GoogleAIStudio announced on November 4 that there are no breaking changes from preview and it is production-ready for browser, mobile and desktop control agents. Official docs list supported models: Gemini 4 Argon, the recommended Gemini 3.8 Flash, Gemini 3.5 Flash-Lite and Gemini 3 Flash Preview. Actions carry an intent field, responses may include a safety_decision requiring user confirmation, and opt-in screenshot prompt-injection detection is available.

Key Takeaways

  • Computer Use is GA on both the Interactions and GenerateContent APIs as of October 30
  • Google says there are no breaking changes from preview, so existing code carries over
  • Three environments: browser, mobile and desktop; Gemini 3.8 Flash is the recommended model
  • Every action includes an intent field; safety_decision can require user confirmation or block
  • Opt-in screenshot prompt-injection detection (Gemini 3.5 Flash-Lite and later, off by default)
Read full story
Linked Models & Agents:
⚡ Gemini 4 Argon⚡ Gemini 3.8 Flash⚡ Gemini 3.5 Flash-Lite🌐 Google DeepMind
#Gemini API#Computer Use#GA#Agents#Google AI Studio#browser automation
A
Anvishaanvisha·
👀112
🔥86
View on 𝕏 ↗

Voyager launches: a desktop agent harness that drives Blender, DaVinci Resolve, After Effects and Ableton — 'the Codex for creative work'

Nullframe (the Moda team, YC F26) shipped Voyager, a local desktop agent harness that drives creative apps through their own scripting APIs, ships free Design/Video/DAW engines, and works with your own Claude/Codex sign-in or API keys. The launch post passed 1.48M impressions.

Voyager launches: a desktop agent harness that drives Blender, DaVinci Resolve, After Effects and Ableton — 'the Codex for creative work'

Key Takeaways

  • Launch post: ~1.48M impressions, 3,074 bookmarks, 2,457 likes on X
  • Drives Photoshop, Illustrator, After Effects, Premiere Pro, DaVinci Resolve, Final Cut Pro, Blender, Unity and Ableton Live; skills also cover Cinema 4D, Houdini, Remotion, Manim
  • ~40 built-in skills, MCP server support, and reuse of skills installed for Claude Code / Codex / Cursor
  • Free app with BYO agent/keys; paid plans from $19/mo, generation billed at provider cost with no markup
  • Apple Silicon + macOS 15 required; Windows 11 / Linux in alpha; desktop v0.1.84 shipped Oct 8
Read full story
Linked Models & Agents:
⚡ Claude Opus 5.5⚡ DeepSeek⚡ Seedance 2.5⚡ Veo 3.1🤖 Voyager🤖 Claude Code🤖 Codex🌐 Nullframe🌐 Moda
#Agent#Creative#Desktop agent#MCP#Blender#Davinci resolve#Yc
O
OpenAI@OpenAI·
👀230
🔥96
View on 𝕏 ↗

Personal desktop agents compared: Grok Bot vs Meta Muse vs OpenAI Dots

After OpenAI launched Dots (GPT-6 Astra, per-dot cloud computer, 4000+ plugins) at DevDay, the personal/desktop agent space has three clear poles: Cursor/xAI Grok Bot (developer teammates + routines), Meta Muse (Secure VM consumer agent; often misheard as “Mate Muse”), and OpenAI Dots (always-on primary Dot). Feature comparison from official docs only—no invented benchmarks.

Personal desktop agents compared: Grok Bot vs Meta Muse vs OpenAI Dots

Key Takeaways

  • “Mate Muse” verifies as Meta Muse (2026-09-08): Meta news + How We Designed Muse; Muse Secure VM + Muse Spark; consumer life/goals agent.
  • OpenAI Dots (DevDay 2026-09-29): Introducing dots; GPT-6 Astra; own cloud computer; 4000+ plugins; ChatGPT Pro / Business Premium; one primary Dot at launch.
  • Grok Bot: Cursor docs — persistent cloud computer, plugins/MCP, skills/routines, multi-Bot; included with paid Cursor plans or SuperGrok / X Premium+ link (Plans).
  • No public three-way official SWE-Bench/Agent Arena bake-off; pick by job: coding/routines/multi-Bot → Grok Bot; consumer life agent → Muse; ChatGPT ecosystem always-on → Dots.
  • Shared pattern: isolated cloud computer + approval for sensitive acts; differ on developer routines/local exec (Grok Bot), WhatsApp/consumer connectors (Muse), plugin breadth + Pro bundling (Dots).
Read full story
Linked Models & Agents:
⚡ GPT-6 Astra⚡ Muse Spark⚡ xAI Grok🤖 Grok Bot🤖 Meta Muse🤖 OpenAI Dots🌐 OpenAI🌐 Meta AI🌐 xAI🌐 Cursor (Anysphere)
#Grok Bot#Meta Muse#OpenAI Dots#Personal Agents#Desktop Agent#Cloud PC#MCP#GPT-6 Astra#Muse Spark#Comparison
ADSponsored
V
Vaibhav (VB) Srivastav@reach_vb·
👀63
🔥91
View on 𝕏 ↗

OpenAI staff: Astra jumps computer-use vs Sol — ScreenSpot-Pro 76.9%→92.7%

OpenAI staffer Vaibhav (@reach_vb) shared GPT-6 Astra computer-use gains versus GPT-5.6 Sol: ScreenSpot-Pro (click targeting) 76.9%→92.7%; OSWorld 2.0 desktop workflows 65.7%→72.6% with 47% less time in latency simulations; Mind2Web web tasks about 1.9× faster with Astra plus an updated Codex harness. Concrete signal for browser/desktop agent and Codex coding workloads.

Key Takeaways

  • ScreenSpot-Pro: 76.9% → 92.7% for click targeting.
  • OSWorld 2.0: 65.7% → 72.6%, with ~47% less time in latency simulations.
  • Mind2Web: ~1.9× faster with Astra + updated Codex harness — relevant for web/desktop agents.
Read full story
Linked Models & Agents:
⚡ GPT-6 Astra⚡ GPT-5.6 Sol🤖 codex🌐 OpenAI
#GPT-6 Astra#computer use#ScreenSpot-Pro#OSWorld#Mind2Web#Codex#OpenAI#GPT-5.6 Sol
M
Meta AI@AIatMeta·
👀29
🔥92
View on 𝕏 ↗

Meta AI Releases Muse Glimmer: 30B Open-Weight Always-On Local Agent Model Under 20GB VRAM

Meta AI released Muse Glimmer, an Apache 2.0 licensed 30B open-weights model engineered for always-on local desktop agents under 20GB VRAM with DFlash speculative acceleration.

Key Takeaways

  • Released under permissive Apache 2.0 license with public weights and developer guides on Hugging Face;
  • Tuned specifically for agent loops to autonomously discover devices, query APIs, and build local web apps;
  • DFlash speculative decoding maintains high-throughput interactivity on consumer-grade Macs and PCs.
Read full story
Linked Models & Agents:
⚡ Llama 3.3 70B
You've caught up with all stories