0xtnp Flow Studio

Queue Google Flow, ChatGPT, Gemini and Grok prompts, attach images, save results, and repair flagged prompts with cloud repair.

0xtnp Flow Studio runs Google Flow, ChatGPT, Grok and Gemini image and video queues from a convenient side panel. • Paste prompts or import TXT, CSV and XLSX files. • Google Flow: Text to Video, Frame to Video, Ingredients to Video, Text to Image, Image to Image and Agent mode. • ChatGPT: generate and save one image per prompt, with optional reference images. • Gemini: Text to Image, Text to Video, Frame to Video and Ingredients to Video. Each prompt gets its own chat, and uploaded images are checked before sending. • Upload start frames and character/location references; @mentions and filenames pick the right images. • Set per-prompt duration, aspect ratio, resolution and outputs, and follow progress live. • Results are saved to Chrome Downloads or a selected project folder; existing outputs are skipped. • Optional Background mode keeps a Flow, ChatGPT or Gemini tab active using local tab sharing (nothing is recorded or uploaded). • Optional Flow Buddy: a small cat mascot that shows queue progress and short messages. It can be turned off in Settings. Prompt repair (optional) When Flow, ChatGPT or Gemini flags a prompt, repair suggests a safer revision that keeps your characters, references, dialogue speakers and timing, validates it, and retries. It runs up to 10 rounds and stops when a manual edit is needed. • Cloud repair: connect once in the Repair tab, no login needed; a free daily allowance applies. • Local repair: an optional Windows companion (Node.js and a signed-in AGY account required). Repair does not guarantee generation or bypass Google's or OpenAI's policies. Permissions Flow Studio uses Chrome's debugger permission to type prompts and click the site's controls the way you would, because these sites ignore simulated clicks. It only acts on the Flow, ChatGPT and Gemini tabs it runs, and only while a prompt is being submitted. Requirements Google Flow access with generation credits; a signed-in ChatGPT account for ChatGPT images; a signed-in Google account with Gemini image and video generation for Gemini. Privacy Images, videos and your queue stay in your browser. Cloud repair sends the flagged prompt text, the site's warning and earlier repair attempts to the repair server and its AI provider; repair records are kept on that server to improve later repairs for your installation. Local repair keeps its history on your PC and sends prompts to the AI provider through AGY. Independent tool; not affiliated with or endorsed by Google or OpenAI. Whats New: 0xFlow 1.13.41 - Storyboard sheets: 4 scenes in 1 image, cut into 4 start frames saved under their own names. Uses far fewer image credits. Works in AI Story and in manual Text to Image / Image to Image queues. - AI upscale on PC: small storyboard frames are sharpened 2x before saving, so their videos look as sharp as from normal images. The upscale model downloads once on first use. - Overnight Factory: run up to 5 AI Story stories in a row, exported as MP4s. - AI Story: new UGC Product Ad template. Upload your product photo, add its real benefits, and AI Story builds a short ad around it. - Template gallery grouped by type, with search. - AI Story: Redo all reference images / Redo all scene images in one click. - Phone browsers: sign in with Google or Discord, Control opens in a tab, and Flow prompts and video saves work on mobile. - New account menu (support, upgrade, log out, quick tour) and a plan card with sign-in at the top of Control. - Auto-Fix: when Flow refuses a scene because of its start image, AI Story makes a new start image and remakes the video by itself. - AI Story: characters keep their screen side between scenes, every line says who the speaker looks at, and children always wear covered clothes. - Smaller download. - Fixed: Agent mode left open in Flow is switched off automatically (fixes "Flow settings trigger not found"). - Fixed: ChatGPT answers with a question instead of an image are retried within seconds; a ChatGPT tab that is still opening no longer stops the queue. - Fixed: storyboard sheets no longer switch to single images by themselves after a reload or a stopped run. 0xFlow 1.13.39 - Gemini: choose the model and thinking level in Settings. Fixed full-size image downloads and small-window detection. - Flow: choose Nano Banana Pro, Nano Banana 2 or Nano Banana 2 Lite for image generation. - Grok: added 15-second videos. - AI Story: new Podcast Hugot Host template with its own host, outfit and studio per project. - Improved Tagalog Drama conversations, scene continuity, character positioning and prop consistency. Ingredients to Video follows character reference images more closely. - Story Workspace: new Cover & caption tools. Manual video queues can also become workspace projects for preview, subtitles and finishing. - Continue to Part 2 after generating videos outside the extension. - Max Speed: switch between three simultaneous Flow prompts and one at a time. - Save reference images with names such as GAB.png for reuse as @GAB. - Auto-Fix: repair flagged dialogue, show changed words in the log and update subtitles accordingly. - Fixes: fewer duplicate Flow submissions, more reliable ChatGPT/Grok result matching, improved scene regeneration and retries, and better handling of brief connection interruptions. - Added a Discord community link in the header. 0xFlow 1.13.38 - AI Story: 10 new templates: Kapre Vlogs, POV History, Pera Matters, Luxury Files, Unsolved PH, Halloween Specials, Sari-Sari Store Confessions, Blotter Ngayon Gabi, Jeepney Dashboard Diaries and Reverse History. - Story Workspace Finish: automatic Reels-style subtitles timed to the voices, hook title, end card, scene transitions, slow zoom, free background music that gets quieter when people talk, subtitles (.srt) and 1080p export (Pro). - Parallel in Flow (beta): send up to 3 prompts at once in the same Flow project. Off by default. - AI Story: people face the person they talk to, and character references are clean (plain background). - AI Story shows a warning when its server cannot be reached. - Fixes: subtitles stay in time even with sound effects, and file names are easier to tell apart. 0xFlow 1.13.37 - Free plan: 20 generations per day. Pro: unlimited generations (see Plans). - Flow: Outputs per Prompt 2-4 now saves every result. - Gemini: stays on your chosen Google account and turns the Images/Videos tool back on by itself. 0xFlow 1.13.36 - AI Story Auto run: references, scene images and videos run by themselves ("Auto: run all steps"), or one step at a time. - New workflow: Ingredients to Video, where videos are made straight from your character and location references (Flow references are tagged into the prompt). - Tagalog Drama: new five-view character sheets on white, so faces and outfits stay the same in every video. - Steadier drama videos: people keep their side, no surprise extra people, outfits stay the same, and stories can move to a second place. - Small progress window on the page during Auto runs. - User Guide icon, and tabs renamed to AI Story and Auto-Fix. - Fixes: Flow stops when credits run out, Text to Image uses your Output count, and Background Mode stays on through Auto. NEW in 1.13.35 • Story Builder: pick a template and get a full story with character cards, start frames and video prompts, ready for the queue. Sign in with Google or Discord; 2 free stories a day, Pro for members of our Discord server. • Grok Image to Image, and picture names in your prompt ("char_01" or "@char_01") are linked to your uploaded pictures in Grok (Image to Image and Ingredients to Video). • Settings → Grok image quality: Speed (default) or Quality 2.0. • Fixed: Grok sometimes made two pictures or videos for one prompt. • Fixed: ChatGPT images are saved right away even when ChatGPT shortens the prompt.