imgs-ai

imgs-ai is a service for generating AI images and short videos from text prompts or reference media. It can be used in a browser workspace or through the imgs-ai CLI and Agent Skills, all sharing one account and credit balance. The site lists a GitHub repository, but the README states that website and CLI source code are not included.

What It Is

imgs-ai is an image and video generation platform with a web workspace plus an Agent-facing CLI and Skill. Users describe a scene or attach reference images, choose a model, resolution and aspect ratio, and receive generated images or MP4 videos. Workflows include text to image, image to image, text to video, image to video, portrait series, video templates and ad-generation tools.

How the CLI and Agent Skill work

The user asks a Skill-enabled Agent to install the CLI and Skill, then approves the account connection in the browser. The Agent checks the credit quote, waits for the user's approval, generates the image or video, and returns the file and task link. The site says this requires a Skill-enabled Agent with command execution, Node.js 20+ and network access. The CLI and Skill support images, text-to-video, image-to-video, optional reference audio and the two-step Hotel Lobby AI template.

Credits, models and results

Image and video generation draw on the same credit balance, and the same rates apply on the web and via the CLI. The site lists multiple image models, including GPT Image 2, several Nano Banana variants, GPT Image 2.5 variants and Grok Imagine Image 2.0, each with its own settings. Completed work appears in generation history. Downloads default to JPG, with PNG and original WebP options. Videos are private and downloadable as MP4. Generated files have a default retention of 15 days.