Invideo
Turns a prompt into a finished video with script, footage and voice
- Made by
- Invideo
- Category
- Video
- Last change
- Oct 28, 2026
Changelog
October 2026
- New feature
Invideo added /sketch, which lets users draw a shot, including framing and camera motion, and have Agent Two render it.
What this means for you: You can draw a shot, including the framing and the camera motion, and Invideo's Agent Two will render it as video. A sketch can describe a shot more clearly than a written prompt.
How to use it: Use the /sketch command in Invideo, draw your shot, and Agent Two renders it.
I make videos in Invideo, which now has a /sketch command where I draw a shot, including framing and camera motion, and Agent Two renders it. Help me plan a sketch for [describe scene]: what framing to draw, how to show the camera motion, and what notes to add. My video is about [topic].
- New model
Invideo added Ideogram Vector 1.0, which generates clean editable SVG logos and vector artwork from a text description or rough sketch.
What this means for you: Invideo can now make a logo from a written description or a rough sketch. You get back an SVG, a vector file that stays sharp at any size and can be edited, with lettering that reads correctly.
Why it matters: It follows Ideogram 4.5 image editing, which Invideo added on Oct 13 and which can edit text inside images.
How to use it: Ideogram Vector 1.0 is live on invideo now. Describe a logo or drop in a rough sketch.
I want to make a logo with Ideogram Vector 1.0 on invideo, which creates an editable SVG from a text description or a rough sketch. Help me write a clear description of the logo, including the text, style and colors. My business is: [describe business and name].
- New feature
Invideo now lets GPT, connected through the invideo MCP, build its own effects such as mixed media and grid effects directly in the editor.
What this means for you: If you connect GPT to Invideo through the invideo MCP (a connection that lets an AI assistant control another app), you can ask it in conversation to build effects like mixed media and grid effects. They appear on a real editing timeline inside the editor.
Why it matters: It builds on the invideo MCP, which launched on Oct 2 so users could connect Claude, GPT or other agents to control the editor.
How to use it: Connect GPT to the invideo MCP, then ask for the effect you want in conversation.
I use Invideo and learned that GPT connected through the invideo MCP can build effects such as mixed media and grid effects on the editing timeline. Help me write clear requests to ask for these effects. My video is: [describe video and the look I want].
- New model
Invideo added Krea 3, with precision edits, generations under 8 seconds and native 2K, unlimited on all annual plans.
What this means for you: Invideo now offers Krea 3, an image model for precise edits, with generations in under 8 seconds and native 2K resolution. It is unlimited on all invideo annual plans.
Why it matters: It comes the day after Kling 4.0 arrived on Invideo on Oct 22, and a week after MiniMax H3 Max on Oct 19.
How to use it: Krea 3 is live on invideo. It is unlimited on all annual plans.
I'm on an invideo annual plan, where Krea 3 is unlimited, with precision edits, generations under 8 seconds and native 2K. Help me plan a batch of images to make and edit with it. My project is: [describe project].
- New model
Invideo added Kling 4.0, offering 30 second clips in one generation, up to 10 keyframes, 15 references and 4K output.
What this means for you: Invideo now lets you make video with the Kling 4.0 model. One generation can produce a 30 second clip, and you can guide it with up to 10 keyframes (specific frames you set) and 15 references (images or clips it should follow). It can output in 4K.
Why it matters: On Oct 1, Invideo launched Agentic Extend, which lengthens shots by up to 30 seconds. Kling 4.0 now gives 30 second clips in a single generation.
How to use it: Kling 4.0 is live in Invideo now.
I make videos in Invideo and just learned Kling 4.0 there can make 30 second clips in one generation, using up to 10 keyframes and 15 references, with 4K output. Help me plan a 30 second clip for [my topic or product]: what to show at each keyframe, and which 15 references I should gather.
- New model
Invideo added MiniMax H3 Max, which can relight scenes, swap characters, convert 3D layouts into photorealistic shots, extend clips and lip sync audio.
What this means for you: Invideo can now use a video model called MiniMax H3 Max. It can relight a scene, swap a character while keeping the same movement, turn a 3D layout into a photorealistic shot, extend a clip, and lip sync audio to any face.
Why it matters: It is the third model Invideo has added in under two weeks, after Nano Banana 2.1 on Oct 8 and Ideogram 4.5 on Oct 13.
How to use it: MiniMax H3 Max is live on Invideo now.
I make short videos in Invideo and just learned it added MiniMax H3 Max, which can relight scenes, swap characters while keeping movement, extend clips and lip sync audio to a face. Help me plan one video for [my topic] that uses two of these abilities, step by step.
- New model
Invideo added the Ideogram 4.5 image editing model, which can change costumes, swap backgrounds, attach objects and edit text inside images.
What this means for you: Ideogram 4.5 is an image editing model now inside Invideo. It can change a costume, swap a background, attach objects to a picture, or edit text directly inside an image.
Why it matters: It follows Invideo's Nano Banana 2.1 addition on Oct 8, which also covered image generation.
I make videos in Invideo, which now has the Ideogram 4.5 image editing model for changing costumes, swapping backgrounds, attaching objects and editing text inside images. Help me write 5 clear editing instructions for my image. My image shows: [describe image] and I want: [describe changes].
- New model
Invideo added Nano Banana 2.1, offering faster image generation, better handling of complex prompts and 4K output.
- Launch
Invideo launched its MCP, letting users connect Claude, GPT or other agents to control the editor and generate images and videos.
- New feature
Invideo launched Agentic Extend in its editor, which extends shots at the head, tail or middle by up to 30 seconds with HDR and 16-bit output.
September 2026
- New feature
Invideo added a "/camera moves" command to Agent Two, letting users pick camera movement presets that the generated clip follows.
- New feature
Invideo launched Effects in its editor, where users describe or reference an effect, get a custom controls panel, and can use 90+ built-in effects.
- New feature
Invideo launched Agentic Sound Design in the invideo editor, where editing agents find or generate music and sound, place it on the timeline, and mix and master it.
- New model
Invideo added the GPT 2.5 Sunburst and Flare image models, supporting storyboards with locked characters, physics shots, type styles and repeated edits without degrading the base.
- Pricing
Invideo made image models GPT 2, Nano Banana Pro and Nano Banana 2 free of credit charges for existing and new annual users in September.
- Launch
Invideo launched the invideo editor, a professional video editor with AI editing agents that build base cuts, add SFX and sound design, and color grade.
Related tools
- HeyGenAI avatar videos and translation
- Luma Dream MachineLuma's AI video and image generator
- SynthesiaAI avatar videos for training and work
- ViduMakes video clips from text, images and reference characters
- LTXLightricks' AI video studio and open video models for turning text or images into scenes.
- OpusClipCuts long videos into short clips for social
