This release marks the largest feature upgrade to date, transforming 120 AI Chat from a text-first client into a comprehensive multimodal AI platform.

Multimodal Rendering

• Rich text and inline images: AI-generated images now render directly within conversation threads

• Unified preview workflow: View, save, and reference images seamlessly without leaving the chat

Expanded AI Model Support

• OpenAI: Full support for the latest GPT models with multimodal and reasoning capabilities

• Anthropic Claude: Complete model family integration from lightweight to advanced reasoning

• xAI Grok: Native support for Grok models, including specialized knowledge domains

• Vercel AI: Optimized models for fast responses and web-development use cases

• OpenRouter integration: Unified access to dozens of AI providers via a single API key

Multi-Provider Image Generation

Integrated support for leading image generation systems

• DALL·E 3

• GPT Image

• Google Gemini 2.0 Flash and Gemini Flash 2.5

• Stable Diffusion

• Grok Image 2

• Hugging Face Flux.1

Users can compare outputs, iterate on prompts, and maintain visual continuity across providers

Reasoning Controls

• Adjustable AI reasoning levels: No, Low, Medium, High

• Balances speed versus depth for casual chats, problem-solving, or advanced research

Prompt Management System

• Prompt library: Organized, searchable storage for reusable prompts

• System prompts: Define AI personality, style, and behavior across conversations

• Full CRUD support: Add, edit, or delete prompt templates within Settings

In-App Guidance

• Tooltip system: Contextual hover guides for faster learning and feature discovery

• Accessibility enhancements: Extra context for UI elements to support diverse workflows

New Specialized Models

• Grok Code Fast 1: Optimized for rapid code generation, debugging, and analysis

• Gemini Flash 2.5 Image Generation: Faster, higher-quality image workflows for prototyping and iteration

Performance and Platform

• Optimized for macOS native performance, ensuring smooth responsiveness

• Cross-platform groundwork underway, preparing expansion to Windows and Linux

New

• Multi-threaded chat: Create multiple threads within a single conversation to interact with different language models side by side, without losing shared context. Ideal for comparing responses, exploring model behavior, or iterating on the same prompt with variations.

• Thread controls:

  • Resize thread panels

  • Edit or delete messages in any thread

First public release of 120 HN, a fully native, cross-platform Hacker News client focused on performance and usability.

Features

• Native macOS implementation

• Optimized rendering pipeline
  Fast UI updates and minimal resource usage across macOS

• Threaded comment viewer
  Efficient rendering of large discussion threads with context preservation.

• AI integration (optional)
  GPT-based summarization and contextual assistance using user-provided OpenAI API key.
  Claude and local model support coming soon.

🔧 What’s New

• Chat Settings Panel:
  Fine-tune how the AI responds with new adjustable parameters:

  • Message Context Limit: Control how many previous messages are included for context in each prompt.

  • Max Output Tokens: Set the maximum length of AI responses.

  • Temperature, Top P, Top K: Adjust randomness and sampling behavior.

  • Presence & Frequency Penalty: Influence repetition and creativity.

• Improved Markdown Rendering:
  Cleaner formatting for code blocks, lists, and text styling to improve readability.

• Rename Chat Title:
  Easily rename your chat sessions for better organization.

• Title Bar Fix:
  Fixed alignment and behavior issues with the window title bar.

• Anthropic Integration:

  • Claude Sonnet 4

  • Claude Opus 4

• Local LLMs supported through Ollama:

  • Llama 3.2, Deepseek R1, and Mistral

120 AI Chat MVP Launch (May 5, 2025)

We are pleased to announce the MVP release of 120 AI Chat on May 5, 2025 on macOS (Universal and Intel chip). This initial release focuses on the core conversation interface with an efficient design, reflecting our emphasis on user experience and conversational intelligence.

🚀 What’s New

• Send and receive messages: Chat seamlessly with AI models using a clean, responsive interface.

• AI-powered replies: Get fast, contextual responses powered by OpenAI’s GPT models.

• Use your own API key: Connect your personal OpenAI API key for full control over usage and billing.

• Dark, Light, and System themes: Switch between themes to suit your environment or system settings.

Stay tuned, more updates are on the way as we expand capabilities and refine the experience.