AI Coding & Agents

swyx

@swyx

@latentspacepod https://t.co/oloSDxf0q7

Sep 23, 2026, 06:43 AM UTC

♡ 18 likes↻ 2 reposts↳ 10 replies

swyx

@swyx

can confirm. ran @latentspacepod AINews side by side with 6 Sol and the difference was night and day: https://t.co/oloSDxf0q7 5.5 Opus is the new default model for AINews going forward. so much more concise and tasteful reporting, with much less slopese than even 5 Opus. https://t.co/P6AXDlBKNP

Sep 23, 2026, 06:43 AM UTC

♡ 219 likes↻ 14 reposts↳ 158 replies

GitHub

@github

Should you review your AI-generated code? Is RAG over? Did Skills make MCP obsolete? The GitHub Podcast responds to these hot takes 👇 https://t.co/7cOBj3MtKW

Sep 23, 2026, 02:30 AM UTC

♡ 96 likes↻ 12 reposts↳ 29 replies

OpenAI Developers

@OpenAIDevs

Thanks to @cerebral_valley for co-hosting our GPT-6 hackathons in SF and NYC 🤝

Sep 22, 2026, 10:19 PM UTC

♡ 33 likes↻ 0 reposts↳ 1 replies

OpenAI Developers

@OpenAIDevs

Get lost in your side projects. One builder at our GPT-6 hackathons brought real star maps and Astra to help you find your way home. https://t.co/D9lkBA7ar4

Sep 22, 2026, 10:19 PM UTC

♡ 615 likes↻ 28 reposts↳ 33 replies

OpenAI Developers

@OpenAIDevs

Tune caching for your application: – Choose which prompt prefixes to reuse with explicit cache breakpoints. – Adjust reasoning effort and tool availability while preserving cached context. – Prewarm shared context so responses start sooner. https://t.co/ILhdJ9uWcZ

Sep 22, 2026, 09:13 PM UTC

♡ 156 likes↻ 8 reposts↳ 6 replies

OpenAI Developers

@OpenAIDevs

We’ve improved prompt caching in the API for GPT-6, helping agents run faster and cost less. Higher cache-hit rates by default mean more input tokens benefit from cached-input discounts of up to 90%. https://t.co/NZse8QKr8G

Sep 22, 2026, 09:13 PM UTC

♡ 1.7K likes↻ 67 reposts↳ 89 replies

OpenAI Developers

@OpenAIDevs

See what’s being reused and what’s preventing cache hits. Track cache-hit rates in the new Prompt Caching Dashboard on https://t.co/9eDbe9PbGF. Use the diagnostics API to identify changes that prevented reuse and estimate how many tokens were affected. https://t.co/M16n4rG1Cl

Sep 22, 2026, 09:13 PM UTC

♡ 152 likes↻ 4 reposts↳ 4 replies

Visual Studio Code

@code

🚀 Claude Opus 5.5 is now available in VS Code with GitHub Copilot. Built for agentic coding, long-running tasks, and knowledge work. Give it a try today! https://t.co/CkQzmZUVPS https://t.co/b66z7s00qs

Sep 22, 2026, 09:08 PM UTC

♡ 827 likes↻ 82 reposts↳ 21 replies

Visual Studio Code

@code

🏃🏽‍♂️‍Agent Merge: Keep Your PR Moving in VS Code Discover Agent Merge (Experimental) in VS Code, a powerful way to keep your pull requests moving after they’re open. Let the agent handle review comments, failed checks, and merge conflicts, apply the necessary fixes, rerun builds and tests, and keep working until your PR is ready to merge. ➡️ https://t.co/9nQZZJ6SlP

Sep 22, 2026, 07:51 PM UTC

♡ 133 likes↻ 14 reposts↳ 14 replies

GitHub

@github

📣 @OpenAIDevs's GPT-6 family is expanding in GitHub Copilot with two additional models now generally available. ☀️ GPT-6 Sol: the balanced model for interactive and agentic coding. 🌕 GPT-6 Luna: a lightweight, cost-efficient model for smaller tasks and the lowest-cost option in the family. Try them in the GitHub Copilot app or @code. https://t.co/gyEGwnYxFT

Sep 22, 2026, 06:41 PM UTC

♡ 403 likes↻ 31 reposts↳ 17 replies

Cognition

@cognition

Try them out today in Devin Desktop and Devin CLI: https://t.co/y3humFMnN0

Sep 22, 2026, 06:22 PM UTC

♡ 20 likes↻ 0 reposts↳ 0 replies

Cognition

@cognition

GPT-6 Sol and Luna are now available in Devin. On FrontierCode 1.1, GPT-6 Sol matches GPT-5.6 Sol’s score at 61% lower cost per task. GPT-6 Luna scores above GPT-5.6 Luna at about a quarter of the cost. At under $0.10 per task, it is the cheapest model on the leaderboard. https://t.co/Mct0ZGxMO2

Sep 22, 2026, 06:22 PM UTC

♡ 751 likes↻ 57 reposts↳ 55 replies

Cognition

@cognition

In our evaluations, we see that GPT-6 Sol reads about 17% less context than GPT-5.6 Sol to reach the same score, and rarely leaves a repo with failing tests after its last edit. GPT-6 Luna’s biggest gains are at low and medium reasoning effort, where it writes real regression tests instead of mock-based ones. Read more: https://t.co/2pz6g4Agux

Sep 22, 2026, 06:22 PM UTC

♡ 53 likes↻ 0 reposts↳ 3 replies

OpenAI Developers

@OpenAIDevs

GPT-6 Sol and GPT-6 Luna are available in the API and rolling out in Codex and ChatGPT Work for Plus, Pro, Business, and Enterprise users. https://t.co/fv2OS3jEI3

Sep 22, 2026, 06:14 PM UTC

♡ 180 likes↻ 8 reposts↳ 5 replies

OpenAI Developers

@OpenAIDevs

GPT-6 Sol and GPT-6 Luna build on Astra’s alignment advances, showing improvements over the GPT-5.6 Sol and GPT-5.6 Luna models in our alignment evaluations. https://t.co/zdJfhce2O0

Sep 22, 2026, 06:14 PM UTC

♡ 174 likes↻ 4 reposts↳ 4 replies

OpenAI Developers

@OpenAIDevs

GPT-6 Sol and Luna just landed in Astra’s orbit. Both launch today with API prices 50% lower than GPT-5.6. Build with Sol. Scale with Luna. To production and beyond. https://t.co/ZCEFp4JdjV

Sep 22, 2026, 06:14 PM UTC

♡ 9.6K likes↻ 750 reposts↳ 277 replies

Cognition

@cognition

On FrontierCode 1.1, our benchmark for real-world engineering tasks that grades mergeability and quality, Opus 5.5 scores 65.3% on Extended — taking the leading spot from Fable 5 at a fraction of the cost. Read more: https://t.co/ELoIAbeica

Sep 22, 2026, 05:35 PM UTC

♡ 21 likes↻ 1 reposts↳ 1 replies

Cognition

@cognition

Try it in Devin Desktop and Devin CLI: https://t.co/y3humFMnN0

Sep 22, 2026, 05:35 PM UTC

♡ 11 likes↻ 0 reposts↳ 1 replies

Cognition

@cognition

Claude Opus 5.5 is now available in Devin. On FrontierCode 1.1, Opus 5.5 takes the #1 sport from Fable 5 at a fraction of the cost. https://t.co/QUhqpGjO9X

Sep 22, 2026, 05:35 PM UTC

♡ 586 likes↻ 35 reposts↳ 44 replies