gabrielmoreira Image Tools
Use these tools when you encounter images you cannot read natively, or when you need to generate new images.
image_read
Send an image to a vision-capable model and get a text description.
Local file (most common):
tai tool image_read --image_path /path/to/image.png --prompt "Describe this image"
URL:
tai tool image_read --image_path https://example.com/photo.jpg --prompt "What is shown?"
With a specific vision provider:
tai tool image_read --image_path /path/to/image.png --prompt "Describe" --provider llm.my-openai:gpt-4o
| Parameter | Type | Required | Description |
|---|---|---|---|
| image_path | string | yes | Image file path or URL |
| prompt | string | no | Analysis instruction (default: describe in detail) |
| max_size | integer | no | Max dimension in pixels for longest edge (default: 1080) |
| provider | string | no | Vision provider connector ID. If omitted, uses default vision model |
Images are automatically resized (preserving aspect ratio) before sending to the vision model. Supported formats: PNG, JPEG, GIF, WebP.
image_generate
Generate a new image from a text prompt (text-to-image). For editing an existing image, use image_edit instead.
Basic usage (always specify output):
tai tool image_generate --prompt "A serene mountain landscape at sunset" --output landscape.png
With specific provider, model and size:
tai tool image_generate --prompt "A futuristic city skyline" --provider llm.my-openai --model gpt-image-1 --dimensions 1792x1024 --output output/city.png
Transparent background (for icons, stickers, product shots):
tai tool image_generate --prompt "A cute fox mascot" --background transparent --output_format png --output fox.png
WebP output with quality control:
tai tool image_generate --prompt "A landscape painting" --output_format webp --quality high --output painting.webp
| Parameter | Type | Required | Description |
|---|---|---|---|
| prompt | string | yes | Text description of the image to generate |
| output | string | yes | Output file path for the generated image |
| provider | string | no | Provider connector ID (use image_providers to list). Auto-selects if omitted |
| dimensions | string | no | Image dimensions (default: 1024x1024). Common: 1024x1024, 1024x1792, 1792x1024 |
| model | string | no | Model name to use. Overrides the provider's default model |
| background | string | no | transparent, opaque, or auto. Use transparent for PNG/WebP with no background |
| output_format | string | no | png, jpeg, or webp. Default: png |
| output_compression | integer | no | Compression level 0-100 for jpeg/webp. Higher = better quality. Default: 100 |
| quality | string | no | low, medium, high, or auto. Higher takes longer. Default: auto |
| extra | JSON | no | Provider-specific parameters as a JSON object (e.g. --extra '{"moderation":"low"}') |
If output is omitted, the image is saved to a default path in the working directory.
image_edit
Edit or transform an existing image based on a text prompt (image-to-image). Use for style transfer, background replacement, adding/removing elements, or any modification that requires a reference image.
Basic usage:
tai tool image_edit --image_path /path/to/photo.png --prompt "Change the background to a beach scene" --output edited.png
With URL image:
tai tool image_edit --image_path https://example.com/photo.jpg --prompt "Make it look like a watercolor painting" --output watercolor.png
With specific provider and model:
tai tool image_edit --image_path /path/to/original.png --prompt "Remove the person in the foreground" --provider llm.my-openai --model gpt-image-1 --dimensions 1024x1024 --output result.png
With mask (edit only the masked region):
tai tool image_edit --image_path /path/to/photo.png --mask /path/to/mask.png --prompt "Replace with a garden" --output edited.png
Transparent background edit:
tai tool image_edit --image_path /path/to/product.png --prompt "Remove background" --background transparent --output_format png --output cutout.png
| Parameter | Type | Required | Description |
|---|---|---|---|
| image_path | string | yes | Reference image file path or URL |
| prompt | string | yes | Text description of the desired edit or transformation |
| output | string | yes | Output file path for the edited image |
| provider | string | no | Provider connector ID (use image_providers with capability=image_editing). Auto-selects if omitted |
| dimensions | string | no | Output dimensions (default: 1024x1024). Common: 1024x1024, 1024x1792, 1792x1024 |
| model | string | no | Model name to use. Overrides the provider's default model |
| mask | string | no | Mask image path/URL. Transparent areas in the mask define the editable region |
| background | string | no | transparent, opaque, or auto. Use transparent for PNG/WebP with no background |
| output_format | string | no | png, jpeg, or webp. Default: png |
| output_compression | integer | no | Compression level 0-100 for jpeg/webp. Higher = better quality. Default: 100 |
| quality | string | no | low, medium, high, or auto. Higher takes longer. Default: auto |
| extra | JSON | no | Provider-specific parameters as a JSON object (e.g. --extra '{"input_fidelity":"high"}') |
If output is omitted, the image is saved to a default path in the working directory.
image_providers
List available image providers filtered by capability.
List image generation providers (default):
tai tool image_providers
List image editing providers:
tai tool image_providers --capability image_editing
List vision (image reading) providers:
tai tool image_providers --capability vision
| Parameter | Type | Required | Description |
|---|---|---|---|
| capability | string | no | image_generation (default), image_editing, or vision |
Returns a list of providers with their available models and connector IDs that can be passed to image_generate, image_edit, or image_read.
Constraints
Use only the parameters listed above for each tool. The supported first-class parameters are: prompt, output, provider, dimensions, model, background, output_format, output_compression, quality, mask (edit only), and extra.
Do not pass n, style, or response_format — they are unsupported and will be ignored or cause errors.
For provider-specific parameters not covered above (e.g. moderation, input_fidelity), pass them through the extra JSON object.