Changelog

GPT-5.4 Suite Now Available

OpenAI's flagship models GPT-5.4 and GPT-5.4 Pro with extended reasoning are now available to all Venice users. Here's everything else we shipped.

New Models

The following models have been added to Venice:

Text Models

• GPT 5.4 — Latest, most advanced text model from OpenAI. Available to all users.

• GPT-5.4 Pro — OpenAI's Pro-tier text model with extended reasoning capabilities. Available to all users.

• Qwen 3.5 35B A3B — Text model from Alibaba Cloud with 35 billion parameters. Now publicly available to all users.

• GLM 4.6 — Text model from Zhipu AI. Now available via API.

• GPT-4o — OpenAI's multimodal text model with 128K context window, supporting text, image, and audio inputs. Available to all users.

Image & Video Models

• Nano Banana 2 — Google’s image generation model with per-unit pricing. Available to all users on a pay-per-use basis

• LTX 2.3 — Video generation model from Lightricks. Available to all users

App

New Features

• Video Prompt Auto-Enhance — Video prompts are now automatically enhanced before generation to improve output quality

• Keyboard Shortcut: Cmd/Ctrl+Enter — Submit prompts using Cmd+Enter (Mac) or Ctrl+Enter (Windows)

• Session Persistence — Chat sessions are now preserved across page reloads and browser restarts

• Route-Based Navigation — Navigation now uses URL-based routing, enabling browser back/forward and direct linking

• Session Selection — Users can now select and switch between active sessions

• Music Generation Queue Polling — Music generation now polls for completion status asynchronously, showing real-time progress updates

• Content Violation UI — Improved messaging and presentation when content moderation flags a request

• Verification Center Header — New header section added to the Verification Center page

• GIF Export — Export generated content as GIF files

• Enter Key Submission — Pressing Enter now submits prompts in applicable input fields

• Logged-In User Redirect — Authenticated users are now automatically redirected away from landing/login pages

• Randomized Video Feed — Videos in the public feed now display in randomized order

• Video Gallery — New gallery view for browsing generated videos

• Video Completion Notification — Users now receive a notification when their video generation finishes

• Recent Chat Images — Recent images from chat history are now accessible in the conversation view

• Variant Model Settings — New settings for configuring model parameters when generating variants

• Venice Banner — New Venice-branded banner added to the interface

• Music Routing and Confirmation — Music generation now includes inline confirmation before submission and updated routing

Wallet and Payments

• Billing Interval Toggle — Monthly checkout now includes a toggle to switch between billing intervals.

• Video Generation Credit Costs — Credit costs are now displayed in the video generator.

• Monthly Crypto Subscriptions — Users can now subscribe to monthly plans using cryptocurrency.

• Credit Cost Reminder Modal — A modal now reminds users of credit costs before confirming an action.

Mobile App

• AI Agent Skills — Agents now support skills, enabling task-specific capabilities inside conversations.

• Multi-Image Support — Send or attach multiple images in a single message.

• Voice Input Reliability — Updated speech-to-text handling for the new Expo SDK, including more reliable auto-submit behavior.

• Android Message Composer — The message input now grows dynamically on Android to avoid typing and layout issues.

• Context Usage UI — Updated the context usage component and fixed layout issues in the context usage popover.

• Conversation Rename Dialog — Added a dedicated dialog for renaming conversations.

• Errored Message Actions — Assistant messages that fail now surface regenerate and delete actions.

• Token Dashboard Navigation — Added authenticated navigation entry points to the token dashboard.

• Settings Navigation — Added Settings to the drawer and refreshed parts of the settings UI.

• Preferred Model Persistence — The model selector now preserves and displays the user's saved model label.

• Credits and Pricing UI — Refreshed credits iconography, price quote styling, and related status colors.

• Variant Settings Navigation — Added a back arrow to variant settings bottom sheets.

• Animated Image Loader — Image generation now shows an animated loading state, with fixes to loader clipping.

• Image Variant Defaults — Updated default handling when selecting image variants.

• Media Item Text Layout — Adjusted line height and removed model labels from media item displays.

• Chat and Search Layout Polish — Refined chat padding and search group header styling.

API

• Music Generation SDK — Added async queue support for music generation in the API SDK

• Auto-Mint API Key — API keys are now automatically generated for new accounts

• API Key Requirement Removed — API key is no longer required for certain operations

• Max Reasoning Effort — New max_reasoning_effort parameter to control reasoning depth on supported models

• Background Removal Model — Replaced BEN-2 with Bria RMBG 2.0 for background removal

• Checkout Session Endpoint — New endpoint for updating an existing checkout session

• Multi-Image Generation — Added support for generating multiple images in a single API request

• Free Tier Moderation — Content moderation now applied to free-tier user requests

• Credit Grant — 1K credits granted to user accounts

• Dynamic Model Selection — Router now dynamically selects the fastest available model for requests

• Annual Subscription Pricing — Annual Stripe subscription price changes from $149 to $180, effective March 1, 2026

• Router Token Limit — Increased the maximum token limit for model router requests

• Grok 4.1 Fast Pricing — Updated per-token pricing for the grok-4.1-fast model

• Stablecoin Payments — Added stablecoin as a payment method

• Music Task Type — Added music as a new task type in the API

Fixes and Improvements

• Improved tooltips now appear after a short delay instead of instantly on hover

• Updated memory behavior in temporary chat sessions

• Improved chat input field now auto-refocuses after sending a message

• Improved border animation on the music generation progress indicator

• Improved moved video models to a different category in the model selector

• Improved video model selector now displays the first 6 models by default

• Updated multimodal model status indicator rendering for new content types

• Improved removed video models selection box from the default view

• Improved upgrade button is no longer shown to users already on a Pro plan

• Improved interface now resets to edit mode after completing an action

• Improved mode switching is now disabled while a generation task is in progress

• Added loading placeholder to reduce layout shift while content loads

• Improved document upload and parsing speed in Memoria

• Added back arrow to variant settings sheets for easier navigation

• Improved adjusted line height and removed model labels from media item text displays

• Updated error message for insufficient account balance

• Improved access by removing the credit requirement for usage

• Improved Phala verification by running it client-side in the browser

• Improved model capabilities to support multi-modal inputs

• Improved rendering of model thinking output in the chat interface

• Improved end-to-end encryption with server key authentication

• Improved privacy capabilities with end-to-end encryption support

• Improved end-to-end encryption with enhanced security measures

• Improved model selector to persist its state across sessions

• Improved certificate security by adding CRL revocation checking

GPT-5.3 Codex Now Available

OpenAI's code-specialized GPT-5.3 Codex model is now available on Venice, optimized for code generation, refactoring, and debugging tasks. Here's everything else we shipped.

New Models

The following models have been added to Venice:

Text Models

• GPT-5.3 Codex — OpenAI's code-specialized text model optimized for code generation, refactoring, and debugging tasks. Available to all users.

Image & Video Models

• Seedream V5 Lite — ByteDance image generation model for text-to-image synthesis. Available to all users

App

New Features

• Batch Image Deletion — Added the ability to select and delete multiple images at once in

• Image-Transition Video on Mobile — Image-transition video models are now available in the mobile app.

• Image Storage Policy — Updated image storage and retention policy.

• Inline Model Selector — Moved the model selector to an inline position within the generation UI.

• Direct Image Generation — Images can now be generated directly from the current context without additional navigation.

• Image Info Button — Added an info button on generated images to view metadata and details.

• Clear Image History — Added option to clear image generation history.

• Image Settings Sync — Image generation settings now sync across sessions and devices.

• Prompt Character Limits — Prompt input fields now display and enforce character limits.

• Prompt Enhancer Info — Added informational tooltip explaining the prompt enhancer feature.

• Rate Limit Display — Added a widget showing current API rate limit usage and remaining quota.

• Enhance Prompt Button — Added a button to automatically enhance/rewrite prompts before generation.

• Pinned Model Selectors — Model selectors now stay pinned to the top of the panel when scrolling.

• Counter Animation — Added smooth animated transitions to numerical counters (e.g., credits, counts).

• Unified Combine UI — Unified the combine feature interface across image and video modes.

Wallet and Payments

• PPU Credit Alert — New dialog warns users about credit costs before regenerating responses from pay-per-use models

• Clickable Credit Elements — Credit indicators in the UI are now clickable

• Credit Balance Display — Credit balance is now visible in the UI

Mobile App

• Mobile Update Dialog — New dialog prompts users to update the app when a newer version is available

API

• Web Scrape & Reasoning — Added web scraping and reasoning capabilities to the API.

• Free Tier Limits Removed — Removed rate limits for the free tier.

• Single Prompt Support — Added support for single-prompt requests (non-chat completions).

• Character-Specific Sampling — LLM sampling parameters now vary per character/persona.

• Negative Prompts Disabled — Negative prompt parameter has been disabled.

• Qwen Image Model — Added Qwen Image model to the v3 inference router.

• Prompt Limit Increase — Maximum prompt length increased to 4,096 tokens.

• Murano API Endpoints — Added new API endpoints for the Murano product.

Fixes and Improvements

• Updated error message text displayed on mobile devices

• Updated image hover action buttons to a circular shape

• Fixed header and footer not appearing when scrolling to the top of the page

• Improved removed the 4:5 aspect ratio from available options

• Improved removed the 4:3 aspect ratio from available options

• Improved the metadata panel shown during variant generation

• Updated the color of the favorite/star icon

• Improved hidden the Variants button when auto mode is active to avoid conflicts

• Improved incompatible video models are now automatically deselected when switching settings

• Fixed settings modal not adjusting its height to fit content

• Improved removed the redundant "New Session" button from the interface

• Improved removed wallet connect button from the video landing page

• Updated default model parameters for Venice models

• Updated default minP sampling parameter for model inference

• Improved handling and response behavior when web search fails

• Improved wallet-based users are now excluded from promotional credit grants

• Fixed image generation prompt not being applied correctly

• Fixed image action buttons not responding to clicks

• Fixed aspect ratio button not functioning correctly in image generation

• Fixed web search and resolution quoting not working correctly

• Improved error handling during backup restore operations

• Fixed style picker and variant selection issues in image generation

• Fixed bugs and improved model selector behavior

• Fixed time-to-first-token timer reporting incorrect values

• Improved document processing reliability

• Improved API to support Anthropic structured output

Claude Sonnet 4.6 Now Available on Venice

Anthropic's latest Claude Sonnet model is now live on Venice. This model brings improved capabilities across reasoning, coding, and conversation at competitive pricing — available to all users.

Here's a selection of everything we shipped over the past weeks.

New Models

The following models have been added to Venice app & API:

New Models

• Claude Opus 4.6 — Anthropic's Opus-tier model with 1M token context window and strong performance across coding, analysis, and complex reasoning. Now available to all users

• Claude Sonnet 4.6 — Anthropic's latest Sonnet with 1M token context, adaptive reasoning, and code performance approaching Opus at Sonnet pricing. Available to all users

• GLM 5 — Zhipu AI's latest-generation text model, now available on Venice via DeepInfra. Available to all users

• GLM 4.7 Flash Heretic — Uncensored, unfiltered text model from Zhipu AI with 128K context window. Available to all users

• MiniMax 2.5 — Text model from MiniMax with multi-turn conversation and strategic reasoning capabilities. Available to all users

Model Updates

• GLM 4.7 — Reasoning mode now enabled, supporting chain-of-thought processing. 128K context window

• Nano Banana Pro — Now includes built-in prompt enhancement for image generation (Pro users)

App

General Updates

• Aspect Ratio Selection — New aspect ratio picker added to image and video generation

• Inpainting Settings — New settings panel with controls for inpainting edits

• Annual Price Increase Banner — Banner notifying users of upcoming annual pricing changes

• Video Studio Link — Added a direct navigation link to Video Studio

• Favorite Models — Save models as favorites for quick access in the model picker

• Image Paste in Video Studio — Paste images directly into Video Studio from your clipboard

• Drag-and-Drop Images — Drag and drop images into Video Studio or chat

• Video Variants — Generate variant versions of existing videos

• Clear All Button — New "Clear All" button to reset current selections

• Multi-Shot Video — Create videos with multiple scenes in a single prompt

• Chat Annotations — Inline annotations now appear alongside chat responses

• Chat Onboarding — Step-by-step walkthrough for new users on first visit to chat

• Video Studio Walkthrough — Guided tour for new users on first visit to Video Studio

• Image Input for Prompt Enhancer — Attach reference images when using the video prompt enhancer

• Category Suggestions — Added category-based suggestions to model or content discovery

• Preferred Model in Settings — Set a default preferred model from the Settings page

• Multimodal Image Loader — Upload images alongside text prompts in multimodal chat

Wallet and Payments

• 3D Secure for Past-Due Payments — Payment retries on past-due invoices now support 3D Secure authentication

• 3D Secure for Subscription Upgrades — Subscription upgrade payments now support 3D Secure authentication

Mobile Web

• Video Studio on Mobile — Video studio link is now visible and accessible on mobile web PWA.

Mobile App

• Video-to-Video Generation — New video-to-video mode allows uploading an existing video as input for generation

• Favorite Models — Star models to add them to a new Favorites section in the model selector

• Favorites Search — Search bar added to the Favorites tab in the model selector

• Memoria — On-device memory system that stores context locally on your device between conversations

• Auto-Generate Memories Toggle — New setting to enable or disable automatic memory generation from conversations

API

• Image Generation Tool — Image generation is now available as a tool in chat completions requests.

• Multi-Prompt Video Inference — Video inference requests now support multiple prompts in a single call.

• Video Inference End Image — New end_image_url parameter for video inference requests to specify a target end frame.

• Image Model Prompt Enhancer Field — Model listing responses now include hasBuiltInPromptEnhancer on image models.

• Image Generation Simple Mode — New simpleMode parameter added to image generation requests.

• Prompt Enhancement Rate Limits — New prompt_enhance rate limit type for prompt enhancement requests.

• Image Request Field — New image field added to the request schema for image-related endpoints.

Fixes and Improvements

• Improved prompt input bar to auto-expand as the user types

• Updated model cards in the model picker to support collapsing

• Improved removed the enhance prompt toggle from the interface

• Added error messages when a response stream fails

• Updated pricing for the PixVerse v5.6 video model

• Updated Pro badges and icons to be visible to signed-out visitors

• Updated Simple mode to hide model name labels

• Updated web search results in chat to support collapsing

• Improved NSFW video filtering in the social feed

• Improved restored the Billing History section in account settings

• Improved redesigned the retry payment flow for past-due subscriptions with clearer status and actions

• Improved wallet reconnection by skipping the Sign-In With Ethereum step for previously authenticated wallets

• Fixed loading dots persisting in auto mode

• Added loading indicators to the Clip library while content loads

• Improved search to include conversations inside folders in results

• Fixed conversation folders section to be scrollable when the list exceeds the visible area

• Improved search to hide folder headers and show only matching conversations

• Added file size error handling for audio and video uploads that exceed the allowed limit

• Fixed video model sub-tabs to be horizontally scrollable in the model selector

• Improved expanded image editing API access to free-tier users

• Increased rate limits for unauthenticated API users

• Updated pricing to tiered model for requests exceeding standard context length

• Updated Opus model requests to use tiered pricing

2 Million Users: Thank You for Building with Venice

Venice has reached 2 million users — a major milestone for our community. Thank you to everyone who has made this journey possible. Here's everything we shipped this week.

New Models

We have added several powerful new models to the Venice platform:

Text Models:

• Kimi K2.5 Model — Offering enhanced performance and capabilities at competitive pricing

• GLM 4.7 Flash Model — Access restrictions lifted, making this powerful model available to all API users

Image & Video Models:

• xAI Grok Imagine — Enhance your creative capabilities with new models for text-to-image, image editing, and video

• Vidu Q3 Video — Publicly accessible video models now available for a wide range of applications and use cases

• Chroma — Now available to all Pro and API users for high-quality, uncensored image generation 

Character Video Generation

Character Video Generation — Create short videos featuring your AI characters. Uses Qwen Edit for image creation and Grok for video and audio production, bringing your characters to life in engaging video format.

• Generate videos with your custom characters

• Integrated image and video/audio pipeline

Venice AI Skill on OpenClaw

Venice AI OpenClauw Skill — Venice is now available as a comprehensive skill on OpenClaw, integrating text, models, embeddings, audio, and web search capabilities. Link:  https://clawhub.ai/jonisjongithub/venice-ai

Venice is now also discoverable on Clawdhub, making it easier to find and use Venice skills.

Combine Images

Merge multiple images into one composite and edit them together with AI-powered blending.

• Multi-Image Selection — Upload and select multiple images to combine into a single composition

• AI-Powered Blending — Intelligent blending that seamlessly merges your images

• Model Selection — Choose from multiple image models including Nano Banana Pro, GPT Image 1.5, Flux 2 Max, Seedream 4.5, and Qwen-Edit

Available now for Pro users and via API

Venice Video Studio

Venice Video Studio — A dedicated workspace for AI video generation. Create videos from text prompts, images, or existing videos with our standout multi-model generation feature that lets you compare results side-by-side.

Key features:

• Generate videos from text prompts, images, or existing videos

• Multi-model generation to compare results across different models

• Side-by-side comparison view

• Available in the App for all users

API Playground

API Playground — Test Venice API endpoints directly in your browser without writing code. After successful alpha testing, the interactive API Playground is now available to all users.

Features:

• Test text, image, and video generation endpoints

• View real-time responses

• Copy code snippets for your integration

App

• Agent Activity — Enhanced workflow with tool calling support for improved user experience

• Cloud SQL Support — Added support for Cloud SQL Socket URLs for Google Cloud SQL users

• Log Enhancements — Improved logging with migration status and modified request and response logs

• Container Architecture — Auto-generated container wrapper structs for improved system architecture

• Model Indicator — New model indicator SVG added for better model representation

• Video Transition — New feature for video transition models with start and end frame specification

• Logo Visibility — Enhanced logo visibility with deliberate design choice

• Image Display — Improved image display in chat with clearer images

• Design Capabilities — Significant enhancement to project's design capabilities with new files and modifications

• Visual Indicators — Added visual indicators for easy issue identification

• Temporary Video Warning — Modified temporary video warning banner for simple mode

• File Upload — Improved file upload with processing indicator and explicit accept attribute

• File Picker — Updated file picker to only allow image files for image editing modes

• Venice Video Studio — Introduced multi-model video generation comparison tool

• Persistent Video Playback — Added persistent video playback in Venice Video Studio

• Video Deletion — Introduced ability to delete individual errored videos from queue

• Translation Updates — Updated translation files for new hint text

• Model Privacy — Integrated existing components to display model's privacy status consistently

• Keyboard Handler — Added keyboard handler to video studio prompt textarea

• Thumbnail Visibility — Improved thumbnail visibility in video studio

• Thumbnail Previews — Introduced thumbnail previews for video queue items

• Video Audio — Introduced dynamic video audio behavior with play on click and mute on hover

• Character Insights — Introduced character insights feature with automatic extraction and new 'Insights' tab

• Widescreen Support — Added support for widescreen 21:9 aspect ratio

• Max Prompt Limit — Introduced max prompt limit for improved user experience

• Chart Component — Created chart component to visualize daily messages

• Video Studio Input — Enhanced overall functionality of Video Studio input panel

• Audio Support — Introduced audio support status display for models in Venice Video Studio

• Browser Notifications — Introduced browser notifications for video completion

• Video Studio Navigation — Modified video studio navigation to always display navigation button

• Chat Inference — Improved chat inference with hidden searching and first token

• Model Selection — Improved model selection process in video studio

• Chat Input Mode — Introduced automatic switching of chat input mode to image mode

• Video Card Width — Updated Venice Video studio UI to enforce minimum video card width

• Recent Images — Introduced feature to quickly select from recent generated images

• Video Generation Time — Introduced feature to log and display video generation time

• Model Selector — Improved model selector with click instead of hover

• Vidu Logo — Added new Vidu video model logo

• Video Generation Retry — Introduced feature to retry character video generation with uncensored model

• Video Direction — Introduced comprehensive video direction for character video generation

• Image Placeholders — Improved image placeholders to appear immediately and render as they complete

• Reasoning Enabled — Introduced 'Reasoning Enabled' toggle in Character Settings modal

• Responsive Layout — Improved responsive layout for consistent user experience

• Character Mode — Introduced new CharacterModeToggle component

• Story Prompts — Introduced feature to provide suggested story prompts for character interactions

• Image Context — Improved user experience by providing essential context for generated images

• Context-Aware Prompt — Introduced setting to disable Context-Aware Prompt Suggestions

• Video Playback — Improved video playback behavior in VideoStudio component

• Refresh Button — Introduced refresh button to conversation suggestions feature

• Prompt Collections — Made prompt collections in Generated Videos section collapsible

• Video Playback Control — Introduced feature to ensure only one video plays at a time

• Chat Output — Enhanced chat output rendering with new components and modifications

• Confirmation Modal — Introduced consistent themed confirmation modal for video deletion actions

• Image Download — Introduced feature to download images after background removal

• Button Visibility — Improved visibility of 'Hide Conversation Suggestions' button

• Search Button — Moved search button next to notifications button and added back button

• Entity Insights — Enhanced entity insights feature with updates to useInsightExtractor.ts hook

• Post Detail Modal — Improved PostDetailModal.tsx file to address scrolling issues

• Story Suggestions — Changed initial value of storySuggestionsEnabled to false

• Action Menu — Updated action menu to remain usable during image and video generation

• Translation Updates — Updated various language files and translations.json file

• Document Upload Button — Updated icon used in DocumentUploadButton component

• Template Columns — Extracted duplicate template columns to a constant and updated panel border styling

• Video Fullscreen — Introduced changes to VideoFullscreenProvider.tsx file and updated translation files

• Recent Images Helper — Introduced feature to display video studio input images in recent images helper

• Edit Image Modal — Introduced ability to paste images directly into Edit Image modal

• Video Gallery — Updated video gallery to maintain 2-column layout until xl breakpoint

• User Experience — Improved user experience with various updates and enhancements

• Responsive Video Gallery — Introduced responsive video gallery grid that adapts to different screen widths

Wallet and Payments:

• Clear Guidance — Understand why the submit button is disabled when watching videos or using credit-billed models without logging in

• Easy Credit Management — Get clear explanations and direct access to purchase more credits when running low in the Video Studio

• Improved Payment Options — Enhanced payment method displays and updated translations for a smoother user experience

• Expanded Access — Free users can now access auto top-up and billing sections for more convenient account management

Performance:

• Persistent Video Generations — Completed video creations are now saved across page refreshes for a seamless editing experience

Mobile App

• Image Sharing — Easily share images directly to the app and attach them to chats

• Responsive Design — Improved layout and responsiveness for a better user experience

• Model Updates — Enhanced model definitions for more accurate and efficient processing

• Technical Updates — Behind-the-scenes improvements for smoother app performance

• App Version Display — Quickly view your app version in the account info section

• Clear Pricing — Subscription prices now clearly displayed in US dollars

• Screen Lock Prevention — Phone screen stays on during tasks like voice recording and AI processing

• Advanced Query — Expanded query functionality for more powerful searching

• Folder Management — Easily create, delete, and manage mobile folders and conversations

• Toast Message Improvement — Update toast messages now fit neatly on mobile screens with reduced font size and text

API

• API Playground — Rolling out to all users after successful alpha testing

• Multi-File Update — Improves overall system performance and functionality by updating multiple files and adding new ones.

• Single File Modification — Enhances user experience by modifying a single file to improve its functionality.

• Chat System Expansion — Expands the features or improves the existing chat system for better user interaction.

• Error Message Improvement — Provides helpful error messages when a CONTENTPOLICYVIOLATION error is encountered, improving user experience.

• Model Access — Allows users to access the model regardless of their subscription tier, with costs being incurred on a pay-per-use basis.

• Debugging Enhancement — Introduces a new error code to differentiate between model runner rejections and validation failures, enhancing debugging and observability.

• Model Metadata Update — Updates the model metadata to provide a direct link to the model's Hugging Face repository.

• Video Inference Capability — Enables video models to utilize start and end frame capabilities.

• Video Model Enhancement — Introduces a new capability to the VideoModelSchema, specifically the supportsEndImage feature.

• Insight Extraction API — Enables the extraction and return of structured profile updates from conversations.

• ImagineArt 1.5 Pro — Enables public access to ImagineArt 1.5 Pro.

• Technical Update — Updates the interface and prioritizes Qwen3 VL 235B for vision requests.

• API Price Sheet Update — Adds ImagineArt15Pro to the API price sheet and updates its FAL analytics URL.

• Model Availability — Demonstrates the ease of managing model availability through configuration updates.

• Request Tracking — Improves the tracking and identification of requests.

• Social Feed Update — Allows users to post images without providing any text fields.

• Security Enhancement — Enhances security by adding a minion authentication header to the TTS request.

• Character Image Generation — Updates the character image generation prompt logic to use the flag.

• Technical Enhancement — Supports enhancements by modifying or adding files.

• Tool Support — Introduces support for xAI server-side tools, specifically websearch and xsearch.

• Whisper Large V3 — Makes Whisper Large V3 public.

• Video Studio Access — Expands access to the video studio, potentially enhancing user experience.

• Variable Addition — Adds a variable to the script and uses it in the function.

• Grok Imagine Update — Updates the Grok Imagine Text-to-Video and Video-to-Video models to reflect their audio support.

• Model Access — Enables public access to the models by setting the configuration.

• Migration Files — Adds new migration files to ensure a smooth transition to the updated schema.

• Code Update — Updates the existing functionality by adding and removing lines of code.

• Cortex Data — Expands the functionality of the API, allowing for more interactions with Cortex data.

• SSE Streaming — Introduces support for nested containers with arbitrarily deep nesting in SSE streaming.

• Model Discovery — Allows API clients to discover and select the model, which was previously hidden.

• Function Update — Updates the function with an added test case to verify the behavior.

• Refresh Index — Introduces a new index to support refresh operations.

• Function Update — Updates the function with additions to the files.

• Technical Update — Includes renaming of queries, optimization of images, and updates to manual header settings.

• Schema Update — Updates the schema.prisma file with notable changes.

• Cortex Data Endpoint — Introduces a new endpoint for listing cortex data, enhancing the API's functionality.

• Cortex Entities — Introduces new API endpoints for managing cortex entities, including retrieval, creation, and publication.

• Image Model Specs — Enhances the image model specs with aspect ratios, providing more detailed information about the images.

• Grok Imagine Defaults — Updates the default resolution to 720p and duration to 10s for Grok Imagine video models.

• Documentation Update — Updates the documentation with comprehensive information.

• Inpaint Model Settings — Introduces aspect ratio settings to the InpaintModelSchema, allowing for more flexible image editing.

• Posts Filtering — Introduces an optional parameter to the posts list endpoint, allowing for server-side filtering of posts by media type.

• Media Type Filtering — Introduces an optional parameter to the posts API client, allowing for server-side media type filtering.

• Model Sets — Introduces the concept of model sets for video models, allowing for better categorization and filtering.

• Video Model Enhancement — Passes the selected video model ID to the prompt enhance endpoint in Video Studio.

• Prompt Enhancement — Enhances prompts for video models by using a video-specific template, including details like motion, camera movements, and audio atmosphere.

• Video Model Schema — Introduces an optional modelSets array field to the VideoModelSchema, allowing video models to be categorized into sets.

• Traffic Routing — Introduces percentage-based traffic routing to LLM host configurations, enabling weighted distribution.

• Payment Retry — Introduces an optional parameter to the past-due invoice payment API, allowing for retrying payments with the customer's current default payment method.

• API Usage CSV — Disables API usage CSV download.

• Subscription Upgrade — Introduces the ability to immediately upgrade Stripe subscriptions.

• Prisma Client Update — Introduces a 30-second statement timeout for PrismaClient.

• Conversations Selector — Introduces a new conversations selector for folder IDs, enhancing the functionality of the useConversations hook.

• API Settings — Introduces an component to the API settings page, enhancing discoverability and user experience.

• RxDB Schemas — Introduces new RxDB schemas for storing audio and video attachment metadata locally.

• Cache Backfilling — Enables the backfilling of missing cache rows for users from USD ledger entries, improving data consistency.

• Sharp Image Processing — Calls Sharp image processing paths to reduce high event loop pauses, particularly for large images over 1MB.

• User Management — Improves the overall functionality of the user management system.

• List Update — Updates the list to include 'Immediate' and adds filters to exclude false positives.

• Entity Insights — Introduces new fields for demographic and physical appearance to entity insights, enhancing the schema.

• Demographic Fields — Adds demographic fields to entity insights, enhancing the functionality of the app's insight extraction capabilities.

• Deprecation Information — Introduces a new flag, apiShowDeprecation, to control the display of deprecation information in API responses.

• Code Cleanup — Cleans up import paths in enhance-image.ts and multi-edit.ts.

• Multi-Image Editing — Introduces a new multi-image editing endpoint, allowing users to submit up to 3 images for editing.

• Routing Enhancement — Allows for more flexible and resilient routing.

• Prisma Configuration — Updates the Prisma configuration to accommodate cleartext connections.

• Billing Endpoint — Introduces a new GET /api/v1/billing/balance endpoint that returns the current balance, daily usage, and epoch reset time, requiring an Admin API key for access.

Fixes and Improvements

Image & Video:

• Fixed issues with image handling in video generation to support start and end images

• Fixed pricing calculations for image-to-video models by adding a fixed input cost field

• Fixed visual effects issues with image-to-video generation, including problems with grey backgrounds

• Fixed image copy protection by adding a custom 'Copy' action that intercepts the context menu

• Fixed video gallery layout to display videos in a consistent 2-column grid

• Improved video studio model matching for better accuracy

• Improved video display quality and playback behavior

• Enhanced video generation guidance with video-specific prompts

• Improved PWA video generation reliability

• Adjusted video player layout for better viewing experience

Chat & UI:

• Fixed post modal scrolling issues

• Improved conversation suggestions visibility

• Improved action menu behavior during inference

• Updated ChatOutputItem renderer for better display

• Improved icon consistency across the app

Wallet & Payments:

• Fixed release and pricing issues

• Improved insufficient credits error handling with clearer messages

• Improved payment method UX with better displays

• Added auto top-up indicator for better visibility

Mobile:

• Improved API page mobile responsiveness

• Improved image handling on mobile

• Improved mobile update toast to fit better on smaller screens

• Prevented screen lock during idle tasks like voice recording

Data & Backend:

• Improved VIS error tracking for better debugging

• Enhanced event loop monitoring to identify performance issues

• Added memory leak prevention rules to streaming handlers

• Added GC pause monitoring for performance optimization

• Improved Redis metrics and lock tracking

• Added 30 second statement timeout for database queries

User Experience:

• Updated video playback behavior for smoother experience

• Made prompt collections collapsible for cleaner interface

• Improved grid layout responsiveness across screen sizes

• Added visual feedback to action buttons

• Prevented blob URL copying for better security

Memoria: Your Private AI Memory

Venice introduces Memoria — a revolutionary local-first memory system that gives Venice the ability to remember context across your conversations while keeping your data completely private.

Unlike cloud-based memory systems, Memoria stores everything directly in your browser using an advanced in-browser vector database powered by FAISS (Fast Library for Approximate Nearest Neighbors). This means:

• Complete Privacy: Your memories never leave your device — no server storage, ever

• Intelligent Recall: Venice can reference past conversations, your preferences, documents you have shared, and important context

• Seamless Experience: Memory works automatically in the background as you chat

• Full Control: Manage your memory documents in the new Chat Memory tab, toggle extraction on/off, and delete anything at any time

• Multi-File Upload: Upload multiple documents to Memory at once

Memoria is now enabled by default for all Pro users. Access your memory settings by clicking the memory icon in chat or visiting the Chat Memory tab in settings.

New Models

We have added several powerful new models to the Venice platform:

Text Models

• Kimi K2.5 — Powerful reasoning model from Moonshot AI with excellent multi-turn conversation capabilities, now available to all users

• Claude Sonnet 4.5 — Anthropic's newest balanced model, available to all users

• Qwen 3 VL 235B — Advanced vision-language model with 235 billion parameters (Pro users)

• Gemini 3 Flash Preview — Google's latest fast inference model, available to all users

• GPT-5.2 Codex — OpenAI's latest coding-focused model (API-only)

Video Models

• LTX V2 — New video generation models with improved quality (all 4 variants available)

• Wan 2.6 Flash — Fast image-to-video generation

Image Models

• ImagineArt 1.5 Pro — High-quality image generation with excellent prompt adherence

• Flux 2 Max — Black Forest Labs' flagship model with exceptional photorealism and detail

• GPT Image 1.5 — OpenAI's advanced image generation with improved text rendering

Model Upgrades

• Veo 3.1 — Now supports 4K resolution output

• Qwen-Edit — Now the default image editing model with improved multi-edit support

• Nano Banana Pro — Now supports 1K, 2K, and 4K resolutions

App

New Features:

• Background Removal — Remove backgrounds from any image with one click (Pro users)

• New Edit Image Models – Now supporting Nano Banana Pro, GPT Image 1.5, Flux 2 Max, Seedream 4.5, and Qwen-Edit

• Combine Images — Merge multiple images together with AI-powered blending, with model selection dropdown

• Transparency Indicator — Checkered background shows transparent areas in images

• Character Selfie Mode — Generate images of your custom characters (Pro users)

• Vision Routing — Vision requests now route to Qwen3 VL 235B in the background

• Character Visualization Button — Quick access to visualize characters

• Import Character Conversations — Bring your character chat history into Venice

• Audio Input for Video — Add audio tracks when generating videos (for supported models)

• Venice Voice Pause — Pause and resume text-to-speech playback

• Web Scrape Toggle — Control URL scraping behavior in Auto mode with smarter URL accessibility checking

• Video Completion Notifications — Browser notifications when your video generation finishes

• Drag-and-Drop Visual Feedback — Clear overlay when dragging files into chat

• Download As Menu — Choose your preferred format when downloading images

• Announcement Toasts — New toast notifications for feature updates

• Past-Due Subscription Visibility — Users can now see and manage past-due subscription status

• Tooltip for Disabled Input — Clear messaging when chat input is disabled and why

• Video Duration Detection — System detects and displays video duration before processing

• Highlight to Chat — Select assistant message content and add it directly to chat input

• Search Provider Switching — Switch between Brave and Google for web search

• Video Start/End Frame — New UI for video transition models

• Auto Prompt Enhancer — Now available in regular image mode, not just simple mode

• Edit Variations — Select 1-4 variations in the edit (inpaint) modal

• Model Variant Tooltip — Tooltip shows which model is selected in variant selector

• Streaming Indicator — Visual … indicator when waiting for LLM chunks

• Social Feed Improvements —  Beautiful new grid layout for browsing community creations

  • Video Autoplay — Videos now autoplay as you scroll through the Social Feed

  • Video Thumbnails — Preview videos directly in the Social Feed before clicking

  • 2-Column Mobile Grid — Better layout for browsing the feed on mobile devices

Wallet and Payments:

• Instant Crypto Payments — Crypto payments now credit instantly to your account

• Wallet Connect Upgrade — Upgraded to latest WalletConnect packages for more seamless sign-in

• Crypto to Credit Card Switch — Crypto subscribers can now switch to credit card billing

• Auto Top-up — Automatically add credits when your balance is low (Pro users)

• Stripe Graceful Degradation — UI gracefully handles when browser blocks Stripe

Performance:

• Sidebar virtualization for faster performance with many conversations

• Improved streaming with pause button and visual feedback

• Context filtering optimization for better KV cache hit rates

• Removed artificial 2.2 second loading wait time

• Improved API page mobile responsiveness

• Cached input tokens now used in web app inference

• Next.js 16 upgrade — builds reduced from 4.5 to 2.5 minutes

Mobile App

• Venice Voice — New text-to-speech functionality that adapts to your language

• Native Markdown Renderer — Better text formatting throughout the app

• Conversations Grouped by Date — Chat history organized chronologically for easier browsing

• Image-to-Video — Turn any image into a video directly from mobile with improved flow

• New Settings Screen — Redesigned settings with better organization

• Image Picker for Combine — Add images to combine directly from your library

• Token Usage Tracking — See your token usage during conversations

• Rate Limit Notice — Clear messaging when you hit rate limits

• iOS Share Extension — Share images directly to Venice.ai from your iOS photo library

• Improved Model Selection UI — New model badges and improved selection interface

• Empty State Improvements — Better handling with keyboard dismissal

• Android APK 1.8.0 — New version available for direct install

API

• New API Dashboard — Track your usage by model and API key with detailed breakdowns

• Nano Banana Pro Resolutions — Now supports 2K and 4K resolutions via API

• SST Endpoint — Speech-to-text endpoint now available for API users

• New Model API Fields — Added privacy, description, betaModel, and deprecation.date to models endpoint

• API Key Prefixes — Keys now prefixed with VENICE-ADMIN-KEY- or VENICE-INFERENCE-KEY- to distinguish types

• API Key Editing — Support for editing existing API keys

• Cache Token Pricing — Optimized pricing for GLM 4.7 and models with input caching

• Claude Cache Write Tokens — Charged at 1.25x input rate

• Search/Scrape Pricing — Flattened to $10/k for all models

• Video Rate Limits — Updated to 40 RPM for API with tier-based limits for UI

• Developer Role Support — Added "developer" role in chat completions schema for GPT-5.2 Codex compatibility

• Video Start/End Frame — Support for video transition models via API

• Insight Extraction API — New endpoint for extracting insights

• Model Suggestions on Error — When a model is not found, similar available models are suggested

Fixes

Wallet & Payments:

• Fixed wallet QR code scanning issues for Base Wallet and Coinbase Wallet

• Fixed WalletConnect signing errors

• Fixed missing fonts in WalletConnect modal

• Improved wallet disconnect consistency

• Fixed duplicate auto top-up charges

• Fixed Pro user chat rate limit bypass

Browser Compatibility:

• Fixed Safari CodeMirror compatibility issues

• Fixed ad-blocker handling for Stripe payments

• Fixed Chrome Chromebook typing issues

Image & Video:

• Fixed image variant preservation during editing

• Fixed image EXIF preservation when downloading

• Fixed character chat model selection errors

• Fixed video upload error handling

• Fixed photo viewer layering issues

• Fixed duplicate share button appearing

• Fixed multi-image switching bug

• Fixed inpaint variations button visibility

• Fixed video credits not showing

• Fixed invalid stored video settings reset

• Fixed image dimension validation in InpaintingModal

Chat & UI:

• Fixed mobile tooltip display issues

• Fixed LaTeX rendering (math expressions like 2^2=4 with citations, double ^^ issues)

• Fixed streaming empty response when reasoning disabled

• Fixed chat messages with line breaks displaying as single line

• Fixed prompt input not clearing after submission

• Fixed "The selected model is temporarily offline" error display

• Fixed hidden close button in video summary

• Fixed divider reasoning spacing

• Fixed text color (text.muted to text.subtle)

• Fixed hover tabs causing premature close in model selector

• Fixed action panel delay and click behavior

• Fixed share conversation for character chats

• Fixed privacy warning flashing and checkbox color

• Fixed chat input when video generation active

• Fixed new chat button not working

• Fixed chat history shared among different chats

• Fixed enhance message undefined check

Data & Backend:

• Fixed database connection errors

• Fixed network timeout handling

• Fixed tag injection for Deepseek 3.2 and Kimi K2

• Fixed billing history transaction time (+5 min offset)

• Fixed Claude assistant message error with tool_calls

• Fixed video quote requirements/schema

• Fixed prompt enhance V3 rewriting issues

• Fixed model router back to Venice Uncensored for proper routing

• Replaced keyv-upstash with @keyv/redis for cache-utils

User Experience:

• Fixed download as image format

• Fixed overflow/viewport height issues (using dvh)

• Fixed default settings route

• Fixed public character cache on moderation

• Fixed burn chart data

• Updated burn history with etherscan and historical fiat prices

• Padded burn history to show minimum 12 months

• Removed exclamation marks on welcome modal

• Updated model selector hover state handling

• Added search icon while searching

• Added missing featured flag for public characters

• Added temporary badge to temporary character chats

• Improved batch deletion UX

Venice Now Hosts All Leading Frontier Models

We've significantly expanded our model library to include the latest and most powerful frontier AI models available. For text, we've added Grok 4.1 Fast with reasoning and vision, Claude Opus 4.5, Gemini 3 Pro Preview, Kimi K2, GPT-5.2, and Deepseek 3.2. For video, we've launched LTX 2 Video, Kling 2.6, and the new Longcat video models. For image generation, GPT-Image-1.5 is now available for alpha users.

These models use credits, and requests through these models are anonymized. Venice is now your one-stop-shop for accessing cutting-edge AI.

App

• Fixed a bug where upscaling or enhancing an image at 1x would switch the conversation type to "image". The conversation type now remains consistent.

• The "Use original prompt" option for upscaling is now correctly capped at the maximum prompt length for the enhance feature.

• Added the ability to mute users directly in the Social Feed.

• The "Auto" model is now hidden from the model switcher when you are in the video category.

• Fixed an issue where credit estimates were not displaying correctly on small screens.

• Added resolution selection to the NanoBananaPro model.

• Added an overflow scroll to the memory documents section to prevent UI issues.

• Moderated posts now appear at the top of the Social Feed for visibility.

• Added a tooltip to the locked label on the token dashboard for better clarity.

• Fixed an overflow issue on the preferences modal for guest users on the profile page.

• Added feedback buttons to the Support Bot responses.

• Hid the mature filter in preferences for non-Pro users.

• Improved system-prompt token calculation for more accurate usage reporting.

• Fixed a bug in multimodal regen.

• Swapped user-facing "Beta" tags to "Alpha" or "Alpha Tester" for consistency.

• Added estimated pricing display for Pay Per Use LLM models in the model selector.

• Added drag-and-drop image upload functionality to the edit, combine, and upscale modals.

• Replaced all instances of "points" with "Venice Credits" in the user profile and settings.

• Relaxed schema validation for chat completions to allow unknown fields, improving compatibility with third-party clients.

• Updated the chat input loading spinner and image loading animation for a smoother experience.

• Added a way to feature characters and updated the default sort on the public characters page to "Featured".

• Fixed duplicate entries on the "Browse Characters" page.

• Fixed the NSFW overlay regeneration logic.

• Updated search provider labels for clarity.

• Added resolution and websearch information to the image details view.

• Fixed style preset settings for models that do not support them.

• Added a message to the Support bot for non-Venice related questions.

• Added a copy button for markdown tables.

• Fixed buttons in the row component from shrinking incorrectly.

Mobile App

• Added 3 new languages (using Google Translate for now).

• Mobile credit purchases now redirect back to the native app after completion.

• Fixed tooltips getting stuck on Android devices.

Models

• Promoted nano banana pro to the beta API.

• Increased the price of Nano Banana Pro.

• Launched the new LTX 2 Video model.

• Added the new Grok 4.1 Fast model to all users and the API, with reasoning and vision enabled.

• Added new Pay Per Use models: Gemini 3 Pro Preview and Kimi K2.

• Upgraded Seedream to version 4.5.

• Launched Kling 2.6 publicly.

• Launched Deepseek 3.2 as a live PRO model, with API support and reasoning enabled.

• Added new Longcat video models.

• Added the new Claude Opus 4.5 model.

• Made GPT-Image-1.5 live for alpha users.

• Removed DeepSeek R1 (beta) from the API.

• Began the two-step process of retiring the Venice-hosted Qwen235B model.

• Added the new GPT-5.2 model.

• Removed Devstral 2 from the API and Kandinsky 5 from the app.

• Enabled reasoning for Deepseek 3.2 again after provider fixes.

API

• Added Fal image generations to the API.

• Updated API documentation for FAL model configurations.

• Added support for the reasoning_effort parameter.

• Added support for the cache_control parameter.

• Added support for prompt_cache_key in the LLM schema to improve cache hits.

• Preserved reasoning_details blocks for advanced tool calling and reasoning.

• Fixed API Dashboard usage display to handle negative values correctly.

• Added support for analysing audio and video files for Gemini 3 Pro/Flash and other compatible models.

• Fixed API KEY consumption, which was incorrectly limiting some users to 50% of their available Diem balance.

• Added consumable balance to video API request headers.

• Added support for Anthropic-format tool configurations and the tool_use/tool_result message format.

• Added cache token tracking and discounted billing for models that support context caching, including Anthropic models.

Token

• Fixed the diem_global_utilization view to include refund entries.

• Busted the Diem usage chart cache to reflect corrected data.

• Added a new "burn" section to the token page.

Fixes and Improvements

• Fixed a bug where prompts longer than the enhance model's limit would cause errors.

• Fixed credit estimates display on small screens.

• Fixed incorrect display of the "Reasoning" tag in text details.

• Fixed an image aspect ratio issue in multimodal chats.

• Fixed resolution examples to use "1K", "2K", and "4K" labels.

• Fixed social user name validation to allow hyphens.

• Fixed FAL image model pricing.

• Fixed regen for character images.

• Fixed aspect ratio issues in selfie mode and the image loader.

• Fixed a bug where the info and feedback buttons were hidden in grid view.

New Pay-Per-Use image models: Nano Banana, Nano Banana Pro, Seedream v4

We launched new pay-per-use image models, available for all users through our new credit system and staking DIEM.

Nano Banana and the new Nano Banana Pro are now live, with increased prompt and token limits. Google's latest image model excels at distilling complex prompts into clear imagery and infographics Nano Banana Pro is truly impressive and we believe it to be a significant step in image models.

Seedream is also now available as a pay-per-use image model.

App

• Released a new, simple model switcher and a new action menu to all users.

• The model switcher now rearranges user tags and triggers settings/actions on hover.

• The action menu fixes a bug with the Video button on img2video models.

• Temporary chats are now live for all users, with the selection persisting across sessions and the state reflected in the chat input.

• Added a progress circle to display context usage and a circular progress indicator with a tooltip for character count.

• Added a confirmation dialog before video regeneration and consolidated confirmation button styling across the app.

• Fixed an issue where the "Create Account" link in the video notice was broken.

• Improved LaTeX rendering to allow more single-dollar-sign inline math correctly.

• Prevented HTML and meta tags from being rendered in chat messages.

• Treats HTML as plain text in user message markdown.

• Fixed a bug with the Support Bot repeating its answer and another with its response rendering.

• Fixed the Character Action Menu button to only open on the intended hover event.

• Updated copy on the delete character confirmation and added validation for chat IDs passed through the URL.

• Disabled the original NanoBanana (V1) model.

• Fixed message/avatar alignment in multi-modal.

• Added support for text parsing from native Word (DOCX) files.

• Added support for XLSX document text parsing using markdown-tables.

• Removed the "new" tag from the support bot.

Models

• GLM 4.6 was set as the default for all users, then reverted due to high output token volume, and later reapplied.

• Added a multi-modal flag to the Gemma model.

• Qwen3-235b is retiring from the in-app model selection, replaced by new Qwen235b API model IDs.

• Added private tags to LLM and image models for better filtering.

• Added anonymized tags to text models.

• Prompt Enhance model has been changed to GLM.

• Wan 2.5 is now the default video model.

• Added negative prompt support to the video settings modal for models that support it.

• Kimi was hidden as reasoning is not supported.

Video

• Set Wan 2.5 as the default video model.

• Changed video model badges layout to prevent the model name from wrapping.

• Added a dialog for credit confirmation before video regeneration.

API

• Added support for reasoning_content instead of <thinking> tags for reasoning models.

• Added mini app tags to the CMS at the request of Coinbase.

Fixes and Improvements

• Fixed an issue with double charging for pay-per-use image models.

• Fixed a missing icon for VEO Fast.

• Fixed an artifact bug and added a watermark to Seedream images.

• Fixed a table overflow-x issue.

• Improved stop/cancel behavior for multi-modal requests.

• Fixed an issue where deleting a folder/character would not delete the conversations within it.

• Memoized the gauth client promise to avoid duplicate promises during await.

GLM 4.6 is now available for Pro users

Developed by Zhipu AI, this model benchmarks extremely high against both closed and open source models. It performs well in character chats and creative writing but mainly excels in tasks where you want a smarter model for analysis or structured problem solving.

Please note that GLM 4.6 is currently live without reasoning.

Web Scraping is Live in the app and API

You can now turn any URL into AI context on Venice Just include a URL in your prompt, and Venice will automatically scrape the page to include as context for your request

Full blog announcement: https://venice.ai/blog/web-scraping-live-on-venice-turn-any-url-into-ai-context

App

• New Library: We are inviting all users to use the all-new Library, a redesigned space to manage and view your generated content.

• Video regeneration: You can now change the model and video settings when re-generating videos, giving you more creative control.

• HEIF Image Support: You can now upload and use HEIF images in the Image Editor and Upscaler.

• Markdown Tables: Added support for GitHub-flavored markdown, including tables, for richer formatting in chat.

• URL Detection: The URL detection feature is now enabled for all users, automatically identifying and processing links in your prompts.

• Video & Image Variations Badge: Added a badge indicator on the chat submit button to show the count of video and image variations to be generated.

• UI/UX Polish:

  • Updated the state label for the web search toggle button for clarity.

  • Fixed an Infura connection error message.

  • Tweaked character escaping to correctly display bold text in edited messages.

  • Fixed text message alignment and edit message padding.

  • Resolved a Support Bot submit button issue on certain screen sizes.

  • Fixed missing action buttons below the last generated video variations.

• Video Preview: Added support for video in the photo preview on the chat interface.

• Mini App: Implemented auto-connect functionality within the mini app on Base.

• Mobile App: Updated the Android APK download URL to point to version 1.3.0.

• ASR on Mobile: Automatic Speech Recognition (ASR) is now available for mobile beta users.

Models

• Added GLM 4.6: GLM 4.6 is now available for Pro users.

• Added Dolphin Mistral 3.2 (Beta): The new Dolphin Mistral 3.2 model is now available in beta.

• Added Veo 3: By popular demand, we have brought back the Veo 3 video generation model.

• Removed Wan 2.5 Safety Check: The safety check on the Wan 2.5 model has been removed for a more open generation experience.

• Decommissioned Flux Models: We have decommissioned the older Flux generation models to streamline our offerings.

Features

• Memoria (Beta): Memoria is now in beta, providing advanced memory retention for longer conversations and more coherent character interactions.

• Character Share Links: Added a share link to the character menu, making it easy to share your custom characters with others.

Fixes & Improvements

• PDF Uploads: Resolved an issue that was blocking multilingual PDFs from being processed correctly.

• Image Upscaling: Improved image quality for upscaled outputs and forced PNG format to remove jpeg compression artifacts.

• Rate Limit Messages: Fixed incorrect rate limit messages for anonymous and free users across chat, image, and video generation.

• Credit Balances: User credit balances now update correctly after usage, DIEM stakes, and unstakes.

API

• Search-Based Pricing: Enabled usage-based pricing for web search for all users.