
What happened
Google Antigravity lead Varun Mohan posted on X that Gemini usage limits were reset for all users and a promised caching fix landed, raising work per limit by 22%.
Why it matters
Existing Antigravity users get more output without paying more, which may make the coding tool more attractive for day-to-day development work.
Summaries like this, every morning.
Antigravity's reset sits alongside a separate update to its /btw command, which now lets users branch a sub-thread for research and promote that sub-thread to the main thread without losing progress on the main work. According to the official account, a demo video was released and users are directed to try it in the CLI.
The same day brought a pricing move from OpenCode, which cut the pay-by-token rates for GLM 5.3 and DeepSeek V4.1 Flash by 60%. Developer dax described the new rates as the formal price rather than a limited-time sale, saying inference runs on OpenCode's own GPUs and is 60% cheaper than Fireworks and Baseten. The new price also applies to direct API calls and to usage beyond a Go subscription's allowance.
Elsewhere in the roundup, Novita AI disclosed that from May 28, 2026 to November 11, 2026 it forwarded some MiniMax model requests to MiniMax's official API instead of processing them in-house. It said a routing misconfiguration meant a forwarding path meant for specific customers was applied to other users, including requests sent via OpenRouter, and that content may have been viewed or stored by a third party. Novita stopped offering the affected models and advised rotating API keys and other secrets included in requests. AppleInsider also reported that TSMC told customers it will stop taking 16nm-and-below chip orders from Chinese companies following tighter US export rules; orders can resume if a US license or exemption is obtained, and no Chinese company appears on the US approved list that includes Apple, AMD and Nvidia.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
A developer's task.md grew to 1,630 lines, about 45% records of finished work, and Claude reported 63 unfinish…

Ziva.sh tested GPT-6 Luna on 10 beginner Godot prompts, with 'Godot 4' specified, no version, and web search;…

A developer's AI-operated cloud browser was blocked with HTTP 403 three times in one day when posting YouTube…

In a 2026-06-19 commit (b5af2be32), Shelly added a flow where a natural-language request is parsed by parseAge…

An FA engineer's Claude Desktop skill post-processes pptx output to remove per-character font direct specifica…

Requests from an iOS app or LINE go to Muse, a Meta personal agent on a cloud VM, which routes jobs through a…
