AI tooling updates and model releases #41

TaeyoungPark

June 8, 2026 — 2 min read

Today's Letter

Meta opens AI app tooling for Portal devices
Google posts Gemma 4 QAT Q4_0 collection

Meta opens AI app tooling for Portal devices

Meta said its AI-powered developer tools that were previously shipped for Meta Quest now also work on Portal and are being opened to all developers
The workflow uses natural-language prompts in an AI coding tool, then deploys the result to Portal with `adb install`
Meta said it built three Portal apps in less than a day, including a to-do app, a podcast player ported from an open-source app, and a home automation dashboard
In Meta's examples, the AI agent handled UI work, local storage, a rendering fix, screen-size adaptation, and smart-home API wiring
The post says the same agentic workflow was introduced for Quest at GDC 2026 and is intended to be hardware-agnostic rather than Portal-specific
Getting started requires enabling ADB on the device under Settings > Debug > ADB Enabled and using Meta's Portal sample app from GitHub
Meta positioned Portal for home and desk use cases such as smart-home controllers, shared message boards, digital art displays, and ambient information panels
Meta's documentation notes there can be impacts on device functionality and account security, and says Portal support is included in its existing CLI and MCP tooling

Source: developers.meta.com
More: this.weekinsecurity.com · news.hada.io · github.com

Google posts Gemma 4 QAT Q4_0 collection

Google published the Gemma 4 QAT Q4_0 collection on Hugging Face.
The lineup lists E2B, E4B, 12B, 26B-A4B, and 31B instruction models.
GGUF variants are available for E2B, E4B, 12B, 26B-A4B, and 31B.
Unquantized QAT Q4_0 models are listed alongside smaller assistant variants.
The collection page shows an update timestamp of 1 day ago, with most model cards updated 2 days ago.
The E4B GGUF card shows an update about 22 hours ago.
Visible download counts on the page range from hundreds to tens of thousands, with 25.1k on the 12B GGUF card.
The collection page shows 64 upvotes at the time captured.

Source: huggingface.co
More: unsloth.ai · zamin.uz

Jocoletter curates AI, software, and product trends for developers and builders.

#Google #Meta

Subscribe to Jocoletter

Anthropic results and NVIDIA training guide #55

Today's Letter 1. Anthropic, Project Fetch Phase Two results published 2. NVIDIA, Low-Precision Transformer Training Guide Anthropic, Project Fetch Phase Two results published * Anthropic published Phase Two results for Project Fetch on June 18, 2026, testing Claude Opus 4.7 on robodog setup and autonomy tasks first run

Agent Infrastructure and Copilot Metrics #54

Today's Letter 1. GitHub adds per-user AI credit metrics to Copilot API 2. AWS, Web Search for Bedrock AgentCore GA GitHub adds per-user AI credit metrics to Copilot API * GitHub added an `ai_credits_used` field to the Copilot usage metrics API for per-user AI credit consumption tracking

Agent deployment, Copilot model shifts, enterprise controls #53

Today's Letter 1. Cloudflare, temporary accounts for agent deployments 2. GitHub Copilot, Opus 4.6 (fast) retirement set for June 29 3. AWS, SageMaker inference metrics on CloudWatch 4. OpenAI adds ChatGPT Enterprise usage analytics and spend controls Cloudflare, temporary accounts for agent deployments * Cloudflare introduced Temporary Cloudflare

Agent Stack, Copilot, and Async Inference #52

Today's Letter 1. GitHub Copilot, context handling and Auto routing update 2. AWS, SageMaker Async Inference adds inline payloads 3. Vercel, Agent Stack and eve unveiled 4. Hugging Face, agentic tooling benchmark for open models published GitHub Copilot, context handling and Auto routing update * GitHub published a June