Dreamina CLI: Bring AI Image and Video Generation into an Agent Workflow
Specific Dreamina CLI Workflow Scenarios
Choose the production situation that matches your automation goal. These prompts cover batch storyboard generation, a scheduled social-video run, and an agent-led asset handoff, so the visuals explain what the workflow is doing rather than showing a vague software concept.
Batch storyboard generation
Create an original 16:9 creative production scene showing a fictional agent turning one campaign brief into six distinct storyboard frames for a short video: wide opening, character close-up, product detail, transition, reaction, and closing shot, organized asset cards, clear visual differences, no readable interface text, no logos, no watermark.
Scheduled social-video run
Create an original 16:9 workflow scene showing a scheduled content job producing a morning social video: a creator brief enters an automated queue, a vertical video preview appears beside a calendar block, a completed result is separated from pending tasks, clean desk lighting, abstract status shapes only, no readable interface text, no logos, no watermark.
Agent asset handoff
Create an original 16:9 post-generation workflow scene where a fictional agent receives three finished AI assets, checks their aspect ratios and thumbnails, marks one image and one video for review, and sends the selected files into an editing timeline; organized folders, clear asset separation, no readable labels, no logos, no watermark.
Explore What You Can Do with Dreamina CLI
Dreamina CLI brings Dreamina’s core generation capabilities into an automatable, agent-ready command-line workflow.
Full-modal generation capabilities
Cover text-to-image, image-to-image, text-to-video, image-to-video, first-and-last-frame video, multi-frame video, multimodal reference video with mixed image, audio, and video inputs, plus image upscaling. One CLI brings Dreamina’s core generation capabilities into the same command-line workflow.
Asynchronous tasks and batch generation
Use a submit-then-query task model: every request returns a submit_id that can be checked later. Add --poll=N to wait automatically after submission, and use generate_num to create 1–10 image variations in one batch.
Agent-native design
Use OAuth device-code login with --headless for non-interactive environments and checklogin --poll to monitor authorization. The installation includes companion Skill files, while a local SQLite database keeps task records available for direct agent orchestration.
Web-aligned models and resolutions
Use Seedream 4.7 and Seedream 5.0 Pro for images, plus Seedance 1.x, 2.0, 2.0 mini, 2.0 fast, and 2.5 for video. Output ranges from 480P to 4K/8K, with --width and --height available when an exact custom size is required.
Discover the Benefits of Using Dreamina CLI
Dreamina CLI is designed for repeatable production: connect generation to code, submit work in batches, and keep every project and task traceable.
Orchestrable and automatable
Turn Dreamina generation into a command-line interface that can connect directly to scripts, CI/CD, test pipelines, and agent workflows. Production teams can trigger repeatable jobs without manually operating the web interface.
Efficient batch work without blocking
Submit multiple images or tasks in one run, save each submit_id, and retrieve results asynchronously. This suits scaled content production because the operator does not need to remain on the page waiting for generation to finish.
Project-based and traceable task management
Use separate Sessions to isolate workspaces by project, while the local tasks.db records generation history and logs. Large batches become easier to manage, review, troubleshoot, and reproduce.
How to Use Dreamina CLI in 3 Steps
Step 1 — Install and log in
Install the CLI, complete device authorization in a browser, and verify the account before generating.
curl -fsSL https://jimeng.jianying.com/cli | bash
dreamina login
# Open verification_uri in a browser and enter user_code
dreamina user_credit
# Success returns credits, user id, and membership levelStep 2 — Submit a generation task
Choose the generation command and pass the prompt, aspect ratio, resolution, duration, or polling interval that matches the job.
# Text to image with a 30-second wait
dreamina text2image --prompt="An orange cat wearing sunglasses" --ratio=1:1 --resolution_type=2k --poll=30
# Text to video
dreamina text2video --prompt="A camera pushes in as an orange cat jumps off a sofa" --duration=5 --ratio=16:9 --video_resolution=720p --poll=30Step 3 — Query and download the result
If the task is still running, keep the submit_id and query it later. Add --download_dir when the completed asset should be saved locally for the next workflow stage.
dreamina query_result --submit_id=YOUR_SUBMIT_ID
dreamina query_result --submit_id=YOUR_SUBMIT_ID --download_dir=./downloadsFAQs about Dreamina CLI
Why does a video generation command tell me to use the web first?
For compliance reasons, the first video generation must be completed in the Dreamina web interface and the authorization confirmation must be accepted there. After that prerequisite, video tasks can be submitted through the CLI; image generation does not have this prerequisite.
What should I do if login says “invalid application”?
Does CLI generation cost money, and how many credits does it use?
Why does the terminal say dreamina: command not found?
What if a submitted task never returns a result?
What information should I send when reporting a problem?
What is a Session, and do I have to create one?
Connect Your Creative Workflow with Dreamina CLI
Plan a structured request, choose the right generation model, and review the returned asset before the next step. When visual creation needs to scale, connect the workflow to dreamina ai video generator and keep each job traceable.