Category: User Interface

Description: Updated the Profile menu dropdown to display the user's credit balance on all pages. Added a "Top up credits" button that links to the Profile page for easy credit management. Partners will see corresponding wallet balance instead.

Impact: Users can now monitor and manage their credit and wallet balance more efficiently, ensuring they stay aware of their available resources.

Category: Functionality

Description: Enabled PDF file uploads for OpenAI models such as o1, gpt-4o, and gpt-4 Turbo, resolving previous upload errors.

Impact: Users can now upload PDFs seamlessly when using supported OpenAI models, enhancing document interaction capabilities.

Category: Functionality

Description: Integrated support for new Gemini models including Gemini 2 Flash, Flash-Lite, and Gemini 2.5 Pro Experimental.

Impact: Users can now leverage these advanced Gemini models to enhance AI prompt chaining capabilities, providing greater flexibility and potential for more complex applications.

Category: Functionality

Description: Updated Perplexity Sonar model to correctly display citations as part of API responses.

Impact: Users gain clearer insights with improved citation handling, although limitations remain in separating "thinking" steps due to model constraints.

Category: Update

Description: Implemented updates to handle deprecation of the max_tokens parameter in OpenAI's API.

Impact: Users will see improved API call compliance and functionality, avoiding potential disruptions due to deprecated parameters.

Category: Functionality

Description: Resolved issues with tokens formula that caused negative max_completion_tokens when prompts were too large.

Impact: Users will experience more reliable token setting functionality, specifically for GPT-4o and other models using max_completion_tokens, enhancing prompt stability and performance.

Category: Update

Description: Updated pricing structure for Perplexity models to reflect new legacy pricing options.

Impact: Users and clients will benefit from clear pricing adjustments, allowing for better budgeting and model use planning.

Category: Functionality

Description: Enhanced Firecrawl model to support mentions of inputs and prompts directly within the URL field, improving usability and integration of inputs with web tools.

Impact: Users can now seamlessly integrate prompt inputs directly in URL fields, streamlining workflows and increasing model flexibility.

CATEGORY: Functionality

DESCRIPTION: Added new Ultra and Raw modes to Flux 1.1 Pro, bringing enhanced performance and versatility to the Flux model.

IMPACT: Users can now access additional model settings that offer improved results and customizable options, expanding the application’s usability.

CATEGORY: Functionality

DESCRIPTION: Updated the Flux model parameter to include an aspect ratio parameter with selectable defaults like Square, Full HD, Standard, and Custom. Custom settings allow preset width and height input.

IMPACT: Users can now better configure image aspects according to their requirements, thereby improving versatility in content presentations.

CATEGORY: User Interface

DESCRIPTION: Resolved a mobile view issue where app descriptions did not wrap correctly when the "With Cover Image" style option was selected.

IMPACT: Enhances readability and presentation of app descriptions on mobile devices.

CATEGORY: Functionality

DESCRIPTION: Integrated support for the newly released GPT-4.5 model from OpenAI.

IMPACT: Users can now utilize the advanced capabilities of GPT-4.5 for improved AI content generation performance.