Faster image model adds GIF generation
🐰 Rabbit takeChatGPT's image model jumps to 2.6, generating images about 50% faster and turning a prompt, photo or sketch into looping GIFs you can speed-adjust and edit frame by frame. Great for memes, social content and brand animations; it's rolling out across all ChatGPT tiers on web, desktop and mobile.
Synthesia plugin brings video creation into ChatGPT
🐰 Rabbit takeThe Synthesia plugin is now available in ChatGPT's plugin menu: describe what you want or drop in a doc or sheet, and it builds a finished video with avatars, voiceover and captions right in the chat, then refines it through conversation. It requires connecting a Synthesia account (a free tier exists), and suits people who want to turn ChatGPT drafts into explainer, training or promo videos.
OpenAI doubles Work and Codex limits, cancels Pro 200 cut
🐰 Rabbit takeAfter subscriber backlash, OpenAI doubled Work and Codex usage limits across Pro 100/200/500 and Business Premium, and scrapped the planned Pro 200 cut. A real win for heavy Codex users, though Plus limits are unchanged. Doubled limits don't mean twice the completed tasks.
ChatGPT adds iCloud Mail and Calendar plugins
🐰 Rabbit takeChatGPT can now connect to iCloud Mail and Calendar to search, draft and send emails and read or update events, on all plans (including free) across web, desktop and iOS, using an Apple app-specific password. Useful for Apple ecosystem users, with the option to revoke access and opt out of model training.
OpenAI mid-tier model speeds up computer use, halves API price
🐰 Rabbit takeSuccessor to GPT-5.6 Terra, built for screen-reading desktop automation like filling forms and logging invoices, with roughly half the error rate and more than double the coding score. Standard API pricing drops to $1 input / $5 output per million tokens—50% lower than before—making it suited to high-volume automation. Available now to ChatGPT Plus/Pro/Enterprise/Edu users in Work and Codex.
Tencent's new Hunyuan Hy4 flagship goes live in Yuanbao with vision and computer use
🐰 Rabbit takeTencent's new Hunyuan flagship scales from under 300B to 770B parameters and, for the first time in the series, can understand images and operate a computer, with stronger long-context and agentic performance. It's already live in Tencent's Yuanbao app, making complex charts, long documents and multi-step tasks more reliable. Tencent says the weights will be open-sourced soon.
Faster, cheaper Kling 4.0 Flash now open to all users
🐰 Rabbit takeKling has opened its faster, cheaper 4.0 Flash model to all users, dropping the Ultra-yearly restriction. It makes 720p clips of 3–20 seconds at about $0.094 per second — roughly 40% cheaper than full Kling 4.0 — and works through Photoshop, Premiere and After Effects plugins. Ideal for high-volume short-form work and quick iteration.
Doubao Work makes everyday Q&A free for individuals
🐰 Rabbit takeFrom November 3, Doubao Work no longer deducts paid quota for individual users' everyday simple Q&A, effectively adding a free lightweight tier; generating documents, making PPTs and deep research may still consume quota, and the system decides what counts as 'simple'. Casual askers can use it more freely, while document/PPT-heavy users see little change.
DeepSeek Harness turns a one-line prompt into a working plugin
🐰 Rabbit takeDeepSeek added Creator Mode to its open-source Harness agent app: describe a tool in plain language and the agent writes, installs and verifies a working plugin — the official demo built a floating Pomodoro timer in about 5.5 minutes. Harness is free and open-source but still a developer preview that runs model-generated code, so it suits tinkerers; everyday users can keep using the DeepSeek web or app chat.
Qwen integrates into Apple Intelligence for China
🐰 Rabbit takeAlibaba and Apple announced Qwen will power Apple Intelligence in China, plugging into Siri, Writing Tools and Visual Intelligence for text/image understanding and generation. Developer beta starts Nov 4 and the public rollout lands with iOS 27.3 in December, so it isn't live for consumers yet. Users opt in, and requests or replies won't be used for training.
