0.5.5 Object + Subject fusion and Character Creator

New Features:
🔄 Object + Subject Fusion
Take a picture of any object and fuse it with a subject image
Perfect for creating custom designs on products
Great for transforming objects into characters/subjects
Combine all sorts of stuff just for fun!

Examples:
• Put character art on t-shirts, mugs, or skateboards
• Transform furniture into character-themed pieces
• Create custom figurines, sculptures, or plush toys

How to Use Object + Subject Fusion:
Click the "Fusion" button
Select "Object + Subject" mode
Upload, paste, drop your object image
Click analyze and use result as is, edit, or reroll
Upload, paste, drop your subject image
Click analyze and use result as is, edit, or reroll
Click the combine button and let AI combine them into a prompt!
Use your prompt as is or edit to your liking!

🎯 Enhanced Default Presets
Improved preset prompts for better, more detailed responses
More focused and specific analysis options
Better structured outputs for image generation

Try it out and share your creations! 🎨✨

🎮 New Feature: Character Creator 🎮

I'm excited to announce a new addition to OllamaVision - the Character Creator! Create detailed characters for your stories, games, or roleplay with your favorite LLM!

✨ Key Features:
• Customizable Fields: Set character attributes including:

Name
Sex
Species (Human, Elf, Dwarf, Android, and many more!)
Setting (Fantasy, Sci-Fi, Cyberpunk, Modern, etc.)
Alignment
Class/Role

🎲 Smart Controls:
• Lock System: Lock any field to preserve your choices while randomizing others
• Random All: Generate random values for all unlocked fields
• Clear All: Quick reset button that respects locked fields
• Editable Fields: Type custom values or choose from presets
• Save Feature: Export your character to a text file

📝 Detailed Output:
• Character Overview
• Personality & Traits
• Physical Description
• Abilities & Skills
• Rich Backstory
• AI Image Generation Prompt

🔄 Workflow Integration:
• Seamlessly works with all supported AI models
• Perfect for worldbuilding and character development
• Generate consistent characters across your preferred settings

Update and try it out now by clicking the "Character Creator" button in the LLM Toys section! 🚀

0.5.1

Introducing Image Fusion and Story Time
🎮 New Feature: LLM Toys Section!

I'm excited to introduce a new "LLM Toys" section in OllamaVision, featuring two powerful creative tools:

🎨 Image Fusion

• Analyze style, subject, and setting separately
• Combines analyses into one cohesive prompt
• Perfect for creating detailed, multi-layered image generation prompts

📚 Story Time

• Transform any image into an engaging story
• Generates detailed narratives with beginning, middle, and end
• Wide-format reading area for comfortable viewing
• Supports drag & drop, paste, or file upload
• Pro tip: Set max tokens to -1 in model settings for best results!

This is just the beginning - more creative LLM toys are in development! Stay tuned for future additions to help unleash your creative potential. 🚀

Compatible with Ollama, OpenAI, and OpenRouter backends

0.4.1

🔄 OllamaVision Update

✨ New Features:
• Redesigned User Prompt System (formerly Response Type)

• More intuitive interface
• Better preset management
• Editable presets that persist between sessions

• New Prompt Prepend System

• Add custom text that will be added before every prompt
• Drag & drop reordering
• Save and manage multiple prepends
• Character limit (1000) to prevent token overflow when combined with user prompt
• Character counter with visual feedback

0.4.0

🚀 OllamaVision Update 0.4.0

✨ UI Improvements
• Added custom branded placeholder image with OllamaVision logo
• Improved button layouts and sizing for better usability
• More polished overall interface

🗜️ New Image Compression Feature
• Added option to compress large images before processing
• Configurable through settings menu
• Helps prevent memory issues and model crashes
• Maintains image quality while reducing size
• Shows compression stats in real-time

💾 Enhanced History System
• Save and manage your image analyses
• View previous results with thumbnails
• Reuse past analyses with original parameters
• Delete individual history items

📝 System Prompt Support
• Added system prompt support for all backends
• Customize model behavior with system instructions
• Works with Ollama, OpenAI, and OpenRouter
• Persists across sessions

🛠️ Enhanced Error Handling
• Better error messages for common issues
• Added request timeout handling
• More informative feedback for model crashes
• Helpful suggestions for fixing common problems

🔧 General Stability
• More robust image processing
• Better memory management
• Improved backend communication reliability

This update focuses on improving the user experience with better history management, system prompt support, and quality-of-life features like image compression to help prevent crashes with large images.

As usual, much more planned, much more coming.

0.3.5

Added easier to see fonts and more importantly, a history view. Will now save your last 20 analysis results including parameter settings for use later. You can delete one or all and if you just want to reuse the image you can drag and drop right into the preview area.

Coming soon: Ability to save history results permanantly.

0.3.4

Another small update to add even more parameters, cleanup lots of code, add a bit of security, and update the way analysis API requests are generated.

0.3.3

Update to include drag and drop image loading capabilities.

0.3.2

Style enhancements to make things easier to see and to better take on style of SwarmUI. Some text can be hard to read on dark themes, I'll be enhancing that at a later time.

Default settings for Temp and Max Tokens has been changed:
Temp from 0.5 to 0.8
Max tokens from 128 to 500

Moved some elements around in modals to allow for more room for future releases.

0.3.0

Includes working remote Ollama connections, model settings for more control over Temp Top P Top K and more, includes OpenRouter API access allowing access to their free models.

0.2.0.0

OllamaVision v0.2.0 Release Notes

I'm excited to announce the release of OllamaVision v0.2.0! 🎉

🚀 New Features & Improvements

🌐 OpenAI Vision Model Integration

You can now connect to OpenAI using your own API key to access their vision models:

1. Go to Settings.
2. Select OpenAI from the dropdown at the top.
3. Enter your OpenAI API Key.
4. Press Save.

You'll be able to utilize any vision models available to you through OpenAI. 🔍🖼️

---

✨ Enhancements

• Better Paste Listener:
  • Now, OllamaVision won’t constantly listen for paste events, preventing loss of generated text and images in the preview area.
  • Simply press the Paste button before pasting to insert your content.
• Performance Optimizations:
  • Small optimizations have been made for smoother performance. ⚡

---

🔧 Removals & Cleanups

• Removed Unnecessary Files:
  • Removed CSS and sortable.min.js files for a leaner codebase. 🧹

---

We hope you enjoy the latest improvements! If you encounter any issues, please feel free to open an issue.

Happy creating! 🚀