Hey Coda Makers!
I’m excited to share the OpenAI Image-1 Pack.
One pack that supports ALL the OpenAI image models:
- DALL-E 2 - Classic and reliable
- DALL-E 3 - Enhanced quality and prompt adherence
- GPT Image 1 - The latest with advanced features like background transparency and custom compression
Key Features:
- Advanced GPT Image 1 Support - Transparent backgrounds, custom output formats (PNG/JPEG/WebP), compression control, and content moderation settings
- Error Handling - Clear error messages when things go wrong
- Base64 to URL Conversion - GPT Image 1 returns base64 data, but the pack automatically converts it to usable URLs via Coda’s TemporaryBlobStorage
- Quality Options - From low-res drafts to HD final images
- Style Control - Vivid or natural styles for DALL-E model
Technical Highlights:
- Model-specific parameter mapping (different models require different parameters, but the pack handles the mapping)
- Automatic quality value translation between models
- Built-in compression validation (0-100% range enforcement)
Important Useage Notes:
- GPT Image 1 requires account verification - Make sure your OpenAI account is verified to access the image-1 model (this took about 10 minutes and required ID verification through a third party)
- Temporary URLs expire after 15 minutes - The pack returns temporary Coda blob URLs that clear automatically. To save images permanently, use:
ModifyRows(thisRow, permanentImageColumn, ParseJSON(tempUrlColumn._Merge()+"","publicUrl"))- this forces Coda to convert the temporary URL to a permanent stored image
What are you building? Would love to see how you integrate AI image generation into your Coda workflows!
Cheers, Jon