OpenAI just released their first open models

View in browser | Past Issue | Subscribe / Unsubscribe

SitePoint Source

Welcome, Developers! 👋

Deployable open-source models on consumer machines are a dying breed but today's release by OpenAI brings major shifts in AI accessibility and huge wins for privacy.

Besides that: real-world Claude Code implementation lessons, Tea app security disaster, modern design systems and much more below.

From our sponsor: Codemetrics

Understand your development team like never before

Codemetrics transforms your team's commits and PRs into actionable insights. Our AI-powered platform reveals who's excelling, who needs support, and how your performance compares to industry standards like DORA metrics.

With developer leaderboards, code reviews, and daily reports, engineering leaders get complete visibility while developers receive well-deserved recognition. Try our free beta with seamless GitHub integration today.

Try for free

🔖 The Reading Room

Articles we have hand-picked for you:

Open models by OpenAI

After a long wait, OpenAI finally released their series of open-weight models that anyone can deploy for free. Open models do have potential for ill use but they decided that far more good than bad will come from it and are excited to see what people build.

The best part is that the smaller model gpt-oss-20b is deployable with consumer GPUs having 16GB of RAM and will enable a lot of use cases that weren't possible due to privacy reasons. These models were tuned for agentic tool usage and perform comparably to frontier models. You can test the models here and easily deploy using any provider like HuggingFace, Ollama etc.

By OpenAI →

6 Weeks of Claude Code

A developer shares how org's internal adoption of Claude Code tackled years of tech all while carrying out regular responsibilities too. It's especially good for eliminating first-step-anxiety but not every org can successfully adopt it. Best results come with monorepos for context, mature boring technologies, and well-structured codebases.

By Orta Therox →

Class Action Lawsuit Against the Viral 'Tea' App

To no one's surprise, there's already a lawsuit against the company behind 'Tea' app for failing to secure users data. Storing users government IDs and selfies (which was supposed to be deleted after verification) in a public Firebase database wasn't enough but they got hit with an actual breach too which caused hackers to download masses of users data including private DMs.

The victims data especially their IDs are now open to be exploited by bad state actors and eventually, more companies storing this verification data will follow since not every org follows strict security standards.

By Lindsey Ellefson →

Coding with LLMs in the summer of 2025

A senior developer, antirez shares specific tips for success with LLMs: Not all LLMs are equal for coding, use Gemini Pro 2.5 and Claude Opus/Sonnet, Avoid "vibe coding" where LLMs work independently, and finally stay in the loop while leveraging AI for bug detection, rapid prototyping, and pair-design activities.

By antirez →

Design Systems in 2025

Leading companies like Google, Apple, IBM, Microsoft, and Adobe prove design systems work at scale. From Material Design's widespread adoption to IBM Carbon's enterprise focus and Apple's attention to detail, these real-world examples demonstrate how robust design systems bridge the gap between development and design teams.

By Atanu Sarkar →

⏳ Back in Time

Most clicks from last newsletter:

🔗 The Link Lounge

Unordered finds from around the web:

Find something cool? You can send us links to feature here via email.

🧰 The Toolbox

Tools and products we're excited about today:

Dyad

Dyad is a local, open-source AI app builder that runs privately on your machine, offering a fast alternative to Lovable, v0, or Bolt with cross-platform support and bring-your-own API keys functionality.

Learn more →

Sidequest

Sidequest is a modern, scalable background job processor for Node.js applications that provides reliable job processing with multiple database backends, worker threads, web dashboard, and comprehensive monitoring capabilities.

Learn more →

EffectTS

Effect is a powerful TypeScript framework providing a functional effect system with rich standard library for building robust applications, featuring type safety, concurrency support, and extensive ecosystem packages.

Learn more →

Ladybird

Ladybird is an independent web browser (no chromium) using a novel engine based on web standards, featuring multi-process architecture with separate renderer, image decoder, and network processes for enhanced security and robustness.

Learn more →

🎤 Your Voice

Your feedback shapes what comes next! We read every email, so simply hit reply and tell us what's on your mind.