Changelog

Text models2

• Venice Small

  Increased context size on Venice Small (Qwen 3 4B) to 40k tokens to fully expose the model’s available context.

  All users

• Venice Large

  Venice Large has been upgraded to Qwen 235B A22B 2507, the latest version from Qwen. The system now programmatically routes thinking and non-thinking requests between two optimized models.

  All users

Image models2

• Anime WAI

  A new anime-focused model trained specifically for high-quality anime-style image generation. It understands anime aesthetics, delivers clean lines and vibrant colors, and produces professional-looking results.

  Pro users

• Qwen Image

  A new image model combining the prompt adherence of Hidream, versatility of Flux, and realism of Venice SD35. Now available for all users.

  All users

Deprecated13

• deepseek-coder-v2-lite
• deepseek-r1-671b
• dolphin-2.9.2-qwen2-72b
• llama-3.1-405b
• qwen-2.5-coder-32b
• qwen-2.5-qwq-32b
• qwen-2.5-vl
• Flux and Flux Uncensored
• flux-dev
• flux-dev-uncensored
• pony-realism
• stable-diffusion-3.5
• Fluently Image Model

Chat10

• Auto Mode

  Auto Mode is now the default for new users. It automatically routes prompts to the best model for answering, eliminating the need to switch between image and text generation manually.

• Current User Time

  Added awareness of current user time to enable more relevant responses for time-sensitive queries.

• Custom System Prompt Indicator

  Added a UI element to indicate when custom system prompts are enabled for the current user.

• File Uploads in Character Chats

  Added support for file uploads in Character chats.

• Bulk Chat Deletion UX

  Improved the bulk chat deletion user experience to prevent accidental deletions during the process.

• Chat Search

  Enhanced chat search to include conversation titles, message content, and attachment names.

• Show 4 more

Image, video and audio6

• Auto-Generate Tags

  Users can now auto-generate tags when posting a new image to the social feed.

• Delete Image Variants

  Users can now delete image variants when generating more than one image at a time.

• Lock Image Settings

  Added support to lock image settings to a specific model for consistent generation.

• Regenerate Any Image

  Any image in the chat history can now be regenerated, not just the most recent one.

• Content Violation Interface

  Improved the content violation interface for image generation to provide more information and allow regeneration of certain prompts using the default model.

• Edit and Enhance Image UI

  Combined the 'Edit Image' and 'Enhance Image' modals into a single interface to simplify image editing workflows.

Account and settings2

• Reverification Enforcement

  Sensitive user actions such as account deletion now require reverification to enhance account security.

• Username Taken Error

  Show an improved error message when users attempt to save a social feed username that is already in use.

Wallet and payments2

• DIEM Launch

  Diem launches on August 20, 2025, with a base mint rate of 90 and adjustment power of 2. Target Diem supply is 38,000, with 10,000 DIEM minted by Venice. VVV inflation reduces from 14M to 10M per year starting August 20.

• Tokenized DIEM

  Launched support for Tokenized DIEM with a new token dashboard design on August 20th, 2025.

General4

• Browser DB Error State

  Added UI states to indicate when the local browser database is inaccessible due to resource constraints imposed by the browser.

• Settings Menu

  A new 'Settings' menu has been added to the left navigation bar to consolidate all Venice settings into one location.

• Thumbs Up/Down Rankings

  Added thumbs up and thumbs down feedback options on results to help Venice better assess model quality based on user input.

• Social Feed Notifications

  Updated the social feed notification drawer to mark all notifications as read when the drawer is viewed.

Mobile app1

• Venice Mobile App

  The Venice mobile app is now available for iOS and Android, bringing private, uncensored AI conversations, image generation, and creative tools to mobile devices.

• API Docs Page

  Added a new API Docs page describing the full Venice model suite available via API.

• API Key Info Route

  Added an API Key Info route to return information about a specific API key. Documentation has been updated accordingly.

• Custom Model Validations

  Implemented custom validations for API models to prevent infinite repetition caused by certain settings on specific models.

• Max Payload Size Limit

  Fixed a bug that was artificially limiting the maximum payload size of API requests, which previously prevented full context utilization.

• Character Creation Modal Layering

  Fixed a layering issue in the Character creation modal where the footer could overlap content or interactive elements. The stacking order has been adjusted to prevent obscuring underlying UI.

  Chat

• Character Modal Footer Layering

  Fixed a layering issue in the Character modal where the footer could overlap content. Adjusted stacking order to improve visibility and clickability.

  Chat

• Conversation Type Reset

  Fixed a bug where creating a new conversation would reset the type back to text. Now carries over the last used conversation type, or selects Auto Mode if applicable.

  Chat

• EXIF Data Stripping

  Fixed a bug that was stripping EXIF data from WebP images upon download.

  Studio

• Original Image Preview

  Fixed a bug that prevented users from previewing the original source image when image mode was enabled.

  Studio