Changelog

We hope you’ve been enjoying your Summer and have found an opportunity to adventure out and explore with family and friends.

Here’s the latest from the Venice team covering the last few weeks of our development. Stay tuned in the coming weeks for a number of major new releases from our team.

Today’s cover art can be found on the Venice Social feed.

Models

• Upgraded Venice Large to Qwen 235B A22B 2507, the latest version of this premier model published by the team at Qwen. Venice now routes thinking and non thinking requests programmatically between the two optimized models. You can learn more about this update here.

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

App

• Launched “Auto Mode” as the default for new Venice users. Auto mode will route prompts to the best model to answer the user’s questions, removing the need for users to swap between image and text generation. You can learn more about this update here.

• Migrated model settings from the top right corner of the chat window into a drop down selector in the chat input. You can learn more about this update here.

• Added a new “Settings” menu to the left nav bar to unify all Venice settings into a single place.

• Added “thumbs up” / “thumbs down” rankings to results to allow Venice to better gauge model quality based on user feedback.

• Permit users to auto-generate tags when posting a new social feed image.

• Added a UI element to show when a user has custom system prompts enabled.

• Added support to lock image settings to a specific model.

• Improved Venice’s models knowledge of its own features and capabilities to permit the model to guide users on how to use Venice more effectively.

• Added the concept of current user time to the models to permit more relevant responses when dealing with time pertinent requests. 

• Combined the “Edit Image” and “Enhance Image” modals into a single user interface for simpler image edits.

• Added UI states to report when the local browser database is inaccessible due to resource constraints imposed by the user’s browser.

• Permit any image in the chat history to be regenerated — previously only the last image in the conversation list would permit regeneration.

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

• Updated content violation interface on image generation to give more information on what happened and provide a mechanism to regenerate certain prompts on our default model.

• Improved bulk chat deletion UX to avoid accidental clicks during flow.

• Permit the deletion of image variants when generating more than 1 image at a time.

• Show an improved error message when users are trying to save a social feed username that’s already in use.

• Fixed a bug where users were unable to preview the original source image when image mode was enabled.

• Fixed a bug that was stripping EXIF data from WebP images when they were downloaded.

• Fixed a bug where when users created a new conversation, the type would be reset back to text. Instead we now carry over whatever the last conversation type they were previously using, or if in auto mode, that is selected.

Characters

• Updated the default character sorting order to require reviews on the latest characters displayed in the list.

• Added the ability for Character chats to support file uploads.

• Fixed a layering issue in the Character modal where the footer could overlap content or interactive elements. The stacking order has been adjusted so the footer no longer obscures underlying UI, improving visibility and clickability.

API

• Added an API Key Info route to return information about a specific API key. API docs have been updated.

• Added an API Docs page describing the Venice model suite.

• Create custom validations for API models to avoid infinite repetition from certain settings on certain models.

• Retired the Fluently image model from the API based on usage patterns.

• Fix API bug that was artificially limiting the max payload size of requests, preventing the full context from being utilized.

It’s been a busy week for our team with 3 major releases, and a slew of updates and bug fixes for you:

• Simple Mode for Pro Users

• Venice… Enhance any Image

• New Image Model: HiDream

  ---

Pro User? Help us test Simple Mode

We're about to roll out the biggest upgrade to the Venice's user experience since our launch: Simple Mode.

Simple Mode provides a streamlined experience by automatically routing your prompts, eliminating the need to manually select a chat type or model (if you ask for an image, it will just know it, without any conception of "Image Mode").

We’ve also added the concept of "image context", which allows you to build images through subsequent prompts. We believe for new users, this will dramatically simplify the user experience, making Venice feel and behave more like other generative AI platforms.

How can you help?

1. Go to your App settings and toggle on Simple Mode. You will see that the model dropdown, and mode selector, will disappear.

2. Chat with Venice and try different prompts in natural language. "Tell me about France" or "Create image of France" etc. Text, web search, image generation, all of it should work automatically.

3. Send your experiences and feedback to [email protected]. Please include screenshots and videos, or share your chat for things that don't seem to work as expected.

We appreciate you taking the time and look forward to hearing your feedback.

Venice… Enhance Image

Pro users can now upscale and enhance any image. Click on the paperclip or drag and drop an image into the image generation chat.

Upscale & Enhance system represents a significant leap beyond traditional pixel-by-pixel methods:

• Use Upscale for accurate enlargements: Use the Upscale feature when you want to maintain the exact style and composition of your image while increasing resolution.

• Use Enhance for creative enhancements: Use the Enhance feature when you want to add details and refinements beyond what's in the original image.

HiDream is now available on Venice

Read more about our latest model in our announcement tweet.

Key Features Prompt adherence - Achieves industry-leading scores on GenEval and DPG benchmarks, outperforming all other open-source models.

Text placement - Effective and accurate text generation and placement, making it useful for graphic design purposes.

Commercial-friendly - Generated images can be freely used for personal projects, scientific research, and other applications.

App

• Added an image generation loading state to make it more clear that the UI is waiting on inference.

• Added a setting to enable display of message timestamps in the App Settings - Featurebase

• Fix crash when editing a chat message and using Space key. - Featurebase

• Updated the error when documents too large to be processed to be human readable.

• Refactored context filtering logic to prevent Your message is too long for the selected model errors when dealing with chat conversations with large histories.

• Ensure that white space submitted in system prompts sent from the app are preserved all the way to the LLM.

• Improved error messages when there are issues communicating with Venice’s inference servers.

• Update the image generation prompt enhancer to ensure messages are constrained to fit within the image prompt limits.

API

• Fixed a regression in streaming responses when invoking tool calling. Added test coverage to avoid future regressions.

• Enabled support for response_format to 405B and Deepseek. Ensure thinking models don't inject <think> tags into the response when response format is enabled.

• Return 504 timeouts when non streaming requests timeout. Recommend use of streaming responses when inference is expected to run for longer durations.

• Remove default `max_tokens` on API calls, permitting inference to use the full context of the LLM unless the user specifies a specific max_tokens

Characters

• Introduced a character avatar popover on larger screens, displaying detailed character information and a quick chat option.

• Added display of character creation date and creator information on public character profiles.

• Introduced clearer visual grouping and dividers in character menus.

• Added a "View & Rate Character" menu item for improved access to rating features

Token

• Updated the token dashboard to fix a few UX issues when used via mobile wallet.

Over the last week, the Venice engineering team has been focused on substantial infrastructure overhaul to our backend inference APIs to improve performance and reliability and support additional scale. Additionally, the team has been working on a comprehensive model curation and testing framework, and preparing to launch a revised image generation infrastructure.

These features will manifest themselves into user visible updates over the coming weeks and we are excited release these updates.

Characters

• Revised our Character moderation system to reduce time to approvals.

Models

• Added support for web search to Llama 3B.

App

• Support horizontal scroll for Mathjax containers on mobile.

• Updated code block and inline code styling for improved readability.

• Fixed a bug that was causing white space to be stripped from uploaded documents prior to LLM processing.