SJinn CLI

Command-line tools for generating images, videos, and audio with SJinn AI.

The npm package installs a small JavaScript wrapper. During postinstall, it downloads the matching prebuilt sjinn binary from the public sjinn-ai/cli GitHub Releases page, verifies its SHA256 checksum, and stores it in the package's vendor/ directory.

Install

npm install -g @sjinn-build/cli

Requirements:

• Node.js 18 or newer
• Network access to npm and GitHub Releases during installation
• macOS, Linux, or Windows on x64/arm64

Authentication

Create an API token from the SJinn web dashboard, then sign in:

sjinn auth login --token <your-token>
sjinn auth whoami

You can also provide the token with SJINN_TOKEN for automation:

SJINN_TOKEN=<your-token> sjinn image generate --prompt "a cat in space"

Image Generation

sjinn image generate --prompt "a cat in space"
sjinn image generate --prompt "landscape" --model nano-banana --aspect 16:9
sjinn image generate --prompt "portrait" --model seedream-v4 --aspect 3:2 --json
sjinn image generate --prompt "logo" --model nano-banana-2 --resolution 4K --async

Video Generation

sjinn video generate --prompt "a rocket launch"
sjinn video generate --prompt "sunset" --model kling3 --duration 5
sjinn video generate --prompt "dancing" --model gemini-omni-video --image-urls "https://example.com/a.png,https://example.com/b.png" --aspect 16:9 --duration 6
sjinn video generate --prompt "timelapse" --model seedance2 --mode fast --async
sjinn video generate --prompt "two characters in a lab" --model seedance2 --media-urls "./character-a.png,./character-b.png,./lab.mp4" --duration 5 --json

Video Composition

sjinn compose --video-urls "./intro.mp4,./outro.mov" --json

Provide at least 2 videos in playback order. Local files support .mp4 and .mov; HTTPS URLs are passed through. Separate audio, voiceover, background music, and transitions are not supported.

Audio Generation

sjinn audio generate --prompt "a calm female voice reading a bedtime story" --model seed-audio-1-0 --async --json
sjinn audio generate --prompt "narrate in the style of @Audio1 with the energy of @Audio2" --audio-urls "./ref-a.mp3,./ref-b.wav" --model seed-audio-1-0 --json
sjinn audio generate --prompt "background music matching the mood of this scene" --image-urls "./scene.png" --model seed-audio-1-0 --async --json
sjinn audio generate --prompt "short upbeat jingle" --model seed-audio-1-0 --download ./jingle.mp3 --json

Task Status

sjinn status <task_id>
sjinn status <task_id> --json

Available Models

Image models:

| Model | Supported parameters | | --- | --- | | gpt-image-2 (default) | --image; --aspect auto, 1:1, 16:9, 9:16, 3:2, 2:3. | | nano-banana | --image; --aspect auto, 1:1, 16:9, 9:16, 4:3, 3:4, 9:21, 21:9, 2:3, 3:2. | | nano-banana-pro | --image; --aspect same as nano-banana; --resolution 1K, 2K. | | nano-banana-2 | --image; --aspect same as nano-banana; --resolution 1K, 2K, 4K. | | seedream | --image; --aspect 16:9, 9:16, 1:1, 4:3, 3:2, 2:3, 3:4. | | seedream-v4 | --image; --aspect 16:9, 9:16, 1:1, 4:3, 3:2, 2:3, 3:4. |

Video models:

| Model | Supported parameters | | --- | --- | | veo3 (default) | --image; --aspect 16:9, 9:16; --end-image for image-to-video. | | gemini-omni-video | --image-urls up to 7 reference images, --video-urls optional reference video (at most 1); --aspect 16:9, 9:16; --duration 4, 6, 8, 10 seconds. | | grok | --image; --aspect 16:9, 9:16, 1:1; --duration 3-15 seconds. | | kling3 | --image; text-to-video --aspect 16:9, 9:16, 1:1; --duration 3-15 seconds; --mode standard, pro; --multi-shot true, false; --end-image for image-to-video. | | seedance2 | --image, --media-urls up to 9 total image/video/audio reference paths or URLs; --aspect 16:9, 9:16, 1:1, 4:3, 3:4; --duration 4-15 seconds; --mode pro, fast; --resolution 480p, 720p, 1080p. |

Seedance 2.0 local --media-urls files support .jpg, .jpeg, .png, .webp, .mp4, .mov, .mp3, .m4a, and .wav. For Seedance 2.0, use either --image or --media-urls; put all multi-reference inputs in --media-urls.

Audio models:

| Model | Supported parameters | | --- | --- | | seed-audio-1-0 (default) | --audio-urls up to 3 reference audio paths or URLs, cited in the prompt as @Audio1, @Audio2, and @Audio3; --image-urls for one reference image. --audio-urls and --image-urls are mutually exclusive. |

Seed Audio 1.0 local reference audio files support .mp3, .m4a, and .wav. Local reference image files support .jpg, .jpeg, .png, and .webp.

Configuration

Config file:

~/.sjinn/config.json

Environment variables:

• SJINN_TOKEN - API token
• SJINN_API_BASE - API base URL, defaults to https://sjinn.ai

Priority: environment variables > config file > defaults.

Manual Binary Download

If npm installation cannot reach GitHub Releases, download the matching asset manually from sjinn-ai/cli releases:

• macOS Apple Silicon: sjinn-vX.Y.Z-darwin-arm64.tar.gz
• macOS Intel: sjinn-vX.Y.Z-darwin-amd64.tar.gz
• Linux x64: sjinn-vX.Y.Z-linux-amd64.tar.gz
• Linux arm64: sjinn-vX.Y.Z-linux-arm64.tar.gz
• Windows x64: sjinn-vX.Y.Z-windows-amd64.zip
• Windows arm64: sjinn-vX.Y.Z-windows-arm64.zip

Use the matching sjinn-vX.Y.Z-checksums.txt file to verify the asset before running it.

Troubleshooting

• Confirm the machine can access https://github.com/sjinn-ai/cli/releases.
• Re-run npm install -g @sjinn-build/cli after transient network failures.
• If the installer reports a checksum mismatch, delete the failed package install and retry.
• For issues, open a ticket in sjinn-ai/cli issues.

License

This package is distributed under the SJinn CLI License. See LICENSE.