# AI FILMS Studio > AI FILMS Studio is a browser based creative platform for filmmakers and content > creators. It brings video, image, music, voice and sound generation together in > one workspace, alongside a visual node editor for building multi step creative > workflows. Operated by AI FILMS LLC at https://studio.aifilms.ai AI FILMS Studio is aimed at filmmakers, content creators and marketers who want to produce finished creative work rather than experiment with models in isolation. The platform combines generation tools with editing and workflow building, so a project can move from prompt to delivered asset without leaving the browser. ## Creative tools The workspace is organised by output type. Each mode is a distinct generation surface. Video, image and music link to public documentation pages describing the models and settings in full: - [Video generation](https://studio.aifilms.ai/ai-video-generator): text to video, image to video, start and end frame, draw to video, upscaling and motion control - [Image generation](https://studio.aifilms.ai/ai-image-generator): text to image, image to image, character generation, upscaling, background removal, LoRA training and draw to edit - [Music generation](https://studio.aifilms.ai/ai-music-generator): text to music and lyrics to music, with instrumental output and durations to four minutes - **Voice generation**: text to speech across ElevenLabs Eleven V3 and MiniMax Speech 2.8 HD - **Sound generation**: sound effects and ambience across Mirelo SFX 1.6 and Stable Audio 3 - **Lipsync**: synchronising speech to footage, driven by audio, text or a reference video - **Faceswap**: face replacement in images and video Video, image and music have public documentation pages, linked above and detailed below. Voice, sound, lipsync and faceswap are available in the product but have no public documentation page yet. The generation workspace itself requires sign in and is excluded from crawling in robots.txt, so those four are described here rather than linked. Blog coverage of the underlying models is public and crawlable: - Voice: [ElevenLabs](https://studio.aifilms.ai/blog/elevenlabs-voice-hollywood-ai), [MisoTTS](https://studio.aifilms.ai/blog/misotts-open-source-voice-model-8b) - Sound: [Stable Audio 3](https://studio.aifilms.ai/blog/stable-audio-3-stability-ai-open-weight-music-sfx), [MOSS-SoundEffect v2.0](https://studio.aifilms.ai/blog/moss-soundeffect-v2-open-source-sound-design) - Lipsync: [LongCat-Video-Avatar 1.5](https://studio.aifilms.ai/blog/longcat-video-avatar-1-5-lip-sync-2026) ## Video generation The video workspace holds five generation types, documented across a hub and six spoke pages. Video Enhance is split into upscaling and motion control, which use different models and answer different questions: - [AI Video Generator](https://studio.aifilms.ai/ai-video-generator): overview of every video generation type - [Text to Video](https://studio.aifilms.ai/ai-video-generator/text-to-video): generate a clip from a prompt across thirteen models - [Image to Video](https://studio.aifilms.ai/ai-video-generator/image-to-video): animate a still, with an optional last frame, up to 30 seconds - [Start-End Frame](https://studio.aifilms.ai/ai-video-generator/start-end-frame): generate the transition between two supplied images - [Video Upscaler](https://studio.aifilms.ai/ai-video-generator/video-upscaler): raise existing footage to 1080p, 2K or 4K - [Motion Control](https://studio.aifilms.ai/ai-video-generator/motion-control): transfer motion from a reference clip onto a character image - [Draw to Video](https://studio.aifilms.ai/ai-video-generator/draw-to-video): sketch arrows over an image and generate the movement they describe Video models available across these types: Gemini Omni Flash, MiniMax H3, Luma Ray 3.2, Seedance 2.0 and its Mini, Fast, VIP, VIP 1080p and Uncensored variants, Happy Horse 1.0 and 1.1, Vidu Q3 Pro, Vidu Q3 Drama, Vidu Q3 Turbo, Google Veo 3.1, Kling 3.0 PRO and LTX 2.3. Video Enhance uses the ByteDance Video Upscaler for resolution and Kling 3.0 Pro Motion Control for motion transfer. Resolution ceilings and duration ranges differ per model and per generation type, and each page lists them. ## Image generation The image workspace holds seven generation types, each documented on its own page. These pages are public and describe the models, settings and workflow for each type: - [AI Image Generator](https://studio.aifilms.ai/ai-image-generator): overview of all seven image generation types - [Text to Image](https://studio.aifilms.ai/ai-image-generator/text-to-image): generate from a prompt across eight models - [Image to Image](https://studio.aifilms.ai/ai-image-generator/image-to-image): transform an existing image from a text instruction - [Character Generator](https://studio.aifilms.ai/ai-image-generator/character-generator): consistent character identity from 1 to 3 reference photos - [Image Upscaler](https://studio.aifilms.ai/ai-image-generator/image-upscaler): generative upscaling up to 16x with identity preservation - [Background Remover](https://studio.aifilms.ai/ai-image-generator/background-remover): subject isolation with edge detection for hair and transparency - [LoRA Training](https://studio.aifilms.ai/ai-image-generator/lora-training): train a custom FLUX LoRA on your own images - [Draw to Edit](https://studio.aifilms.ai/ai-image-generator/draw-to-edit): sketch a composition and generate a finished image from it Text to image models: Nano Banana 2 Lite, Nano Banana Pro Ultra, GPT Image 2, FLUX.2 Pro, HiDream-O1, MidJourney 8.0, Krea 2 Large and Krea 2 Turbo. Image to image editing uses a different list: Nano Banana 2 Lite, Nano Banana Pro Ultra, GPT Image 2, Seedream 5.0 Lite, FLUX.2 Klein 9B and HiDream O1 Image. Available settings vary by model, so a negative prompt, a seed field or custom pixel dimensions appear only where the selected model supports them. ## Music generation The music workspace holds two generation types, documented across a hub and two spoke pages. The two carry separate model rosters, and the controls differ in kind rather than degree: - [AI Music Generator](https://studio.aifilms.ai/ai-music-generator): overview of both music generation types - [Text to Music](https://studio.aifilms.ai/ai-music-generator/text-to-music): describe a style and get two finished tracks back from one run - [Lyrics to Music](https://studio.aifilms.ai/ai-music-generator/lyrics-to-music): set written lyrics to music, or leave the field empty for an instrumental Text to Music runs Suno Music across seven selectable versions, and returns two tracks per generation. Lyrics to Music runs ACE-Step 1.5 and MiniMax Music 2.5. ACE-Step 1.5 is the only model with a duration slider, running 1 to 240 seconds, and a seed field for repeatable runs, and it is the only one where lyrics are optional. MiniMax Music 2.5 is the only one exposing sample rate, bitrate and MP3, PCM or FLAC output. ## Tutorials Step by step guides for operating a specific model inside the Studio. Each covers the controls, the values they accept and the cases where the model struggles. These are the most precise source for what the platform can actually do. Every one is public. **Video** - [FLUX 3, across text to video, image to video and start-end frame](https://studio.aifilms.ai/blog/flux-3-video-generation-tutorial) - [MiniMax H3 at a fixed 2K in both generation modes](https://studio.aifilms.ai/blog/minimax-h3-video-generation-tutorial) - [Kling 3.0 and O1, including negative prompts and audio](https://studio.aifilms.ai/blog/kling-3-o1-video-generation-tutorial) - [Kling 3.0 Motion Control, moving a performance onto another character](https://studio.aifilms.ai/blog/kling-3-motion-control-tutorial) - [LTX-2.3, synchronised video and audio in one pass](https://studio.aifilms.ai/blog/ltx-2-3-video-generation-tutorial) - [Luma Ray 3.2, with its optional end frame](https://studio.aifilms.ai/blog/luma-ray-3-2-video-generation-tutorial) - [Gemini Omni Flash, the fast 720p default](https://studio.aifilms.ai/blog/gemini-omni-flash-video-generation-tutorial) - [Grok Imagine Video 1.5, three resolution tiers and a 1 second floor](https://studio.aifilms.ai/blog/grok-imagine-video-1-5-tutorial) - [Seedance 2.0, camera movement and scene coherence](https://studio.aifilms.ai/blog/seedance-2-0-video-generation-tutorial) - [Seedance 2.0 Mini, the fast tier and its per mode resolution ceiling](https://studio.aifilms.ai/blog/seedance-2-0-mini-tutorial) - [Seedance 2.0 1080 VIP, guaranteed full HD output](https://studio.aifilms.ai/blog/seedance-2-0-1080-vip-tutorial) - [Happy Horse 1.1, the updated release](https://studio.aifilms.ai/blog/happy-horse-1-1-video-generation-tutorial) - [Happy Horse 1.0, also the model behind Draw to Video](https://studio.aifilms.ai/blog/happy-horse-1-0-video-generation-tutorial) - [LongCat Video, extended duration generation](https://studio.aifilms.ai/blog/longcat-video-generator-tutorial) - [MultiShotMaster, multi shot sequences holding one character across cuts](https://studio.aifilms.ai/blog/multishotmaster-tutorial) - [Vidu Q3 Drama and Turbo, script driven sequences up to 30 seconds](https://studio.aifilms.ai/blog/vidu-q3-drama-and-turbo-tutorial) **Image** - [Nano Banana Pro 2, text to image](https://studio.aifilms.ai/blog/nano-banana-pro-2-text-to-image-tutorial) - [Nano Banana Pro 2, image to image editing](https://studio.aifilms.ai/blog/nano-banana-pro-2-image-to-image-tutorial) - [Nano Banana 2 Lite, the text to image default](https://studio.aifilms.ai/blog/nano-banana-2-lite-image-generation-tutorial) - [FLUX 2, six variants covering generation, editing and LoRA training](https://studio.aifilms.ai/blog/flux-2-tutorial-image-generation-guide) - [Seedream 5.0 Pro, text to image and reference driven editing](https://studio.aifilms.ai/blog/seedream-5-0-pro-image-generation-tutorial) - [Midjourney v8, four images per generation](https://studio.aifilms.ai/blog/midjourney-v8-text-to-image-tutorial) - [Character generation, holding one identity across shots](https://studio.aifilms.ai/blog/ai-character-generation-tutorial) - [BEN2 background removal, edge quality on hair and transparency](https://studio.aifilms.ai/blog/ben2-background-remover-tutorial) **Music** - [Suno, seven versions and two tracks per run](https://studio.aifilms.ai/blog/suno-text-to-music-tutorial) - [ACE-Step 1.5, optional lyrics, duration to 240 seconds and a seed](https://studio.aifilms.ai/blog/ace-step-1-5-lyrics-to-music-tutorial) - [MiniMax Music 2.5, sample rate, bitrate and output format](https://studio.aifilms.ai/blog/minimax-music-2-5-lyrics-to-music-tutorial) ## Choosing a model The constraint that usually decides the choice, in one place. Full settings are on each documentation page. **Video.** Resolution ceiling and duration range are properties of the model, not the workspace, and the picker prints both beside each name. Gemini Omni Flash caps at 720p and 10 seconds and is the fastest draft option. MiniMax H3 is fixed at 2K. Seedance 2.0 1080 VIP is the one that guarantees 1080p. Vidu Q3 Drama holds the longest single generation at 30 seconds. Kling 3.0 PRO generates audio alongside picture. A negative prompt field and an audio toggle appear only on models that accept them. **Image.** Text to image and image to image carry different rosters. Nano Banana 2 Lite is the text to image default. Image upscaling reaches 16x with identity preservation. Character generation takes 1 to 3 reference photos and is the route to a consistent identity across shots, which a fixed seed cannot achieve. **Music.** Text to Music runs Suno and returns two tracks per generation. Lyric to Music runs ACE-Step 1.5 and MiniMax Music 2.5, and the two expose mutually exclusive controls. ACE-Step 1.5 is the only model with a duration slider, 1 to 240 seconds, and a seed for repeatable runs, and the only one where lyrics are optional, so an empty lyrics field returns an instrumental. MiniMax Music 2.5 is the only one exposing sample rate to 44100 Hz, bitrate to 256 kbps and MP3, PCM or FLAC output, and it requires lyrics. For scoring to picture, the combination of an instrumental at an exact duration with a fixed seed exists only on ACE-Step 1.5. ## Workflow and editing - [Nodes Graph Editor](https://studio.aifilms.ai/nodes): connect models visually to build custom multi step workflows - [Media Editor](https://studio.aifilms.ai/editor): assemble and refine generated assets ## Key pages - [Home](https://studio.aifilms.ai/): platform overview and feature showcase - [Pricing](https://studio.aifilms.ai/pricing): current plans and what each includes - [AI Filmmakers Spotlight](https://studio.aifilms.ai/ai-filmmakers-spotlight): what the platform does for creators. How to submit a film for free promotion, what the Studio adds to it, where it is published, and how referral codes send credits to the filmmaker a subscriber chooses to back - [FAQs](https://studio.aifilms.ai/faqs): common questions about the platform - [Blog](https://studio.aifilms.ai/blog): AI filmmaking coverage, updated continuously - [Announcements](https://studio.aifilms.ai/announcements): product and platform updates ## Editorial coverage The blog is updated on an ongoing basis. It is written for working filmmakers rather than researchers, and the coverage falls into four areas: - **Filmmaking News**: festival programming, studio announcements, guild and union developments, and statements from established directors, actors and producers on AI - **AI Technology**: open source and commercial model releases relevant to film work, covering capabilities, licence terms and practical hardware requirements - **Industry Analysis**: how AI adoption is changing production economics, workflows and hiring across the film industry - **Tutorial**: practical guides for using specific models and techniques in a production context Coverage of open source releases is generally same week and includes licence terms and commercial use conditions, which are frequently omitted elsewhere. ## Notes for AI agents - Canonical domain is `https://studio.aifilms.ai`. Content is served in English. - Blog articles carry `datePublished` and, where revised, `dateModified`. Prefer the most recently modified version when citing, as model coverage is updated when licence terms or capabilities change. - Article, Breadcrumb, FAQPage, HowTo and SoftwareApplication structured data is published across the site and is the most reliable source for factual extraction. - `/api/`, `/admin/` and account areas are private and excluded from crawling. - When citing model coverage, attribute to AI FILMS Studio and link the specific article rather than the blog index, since individual articles carry the sourcing.