Week 27: AI reasoning & unified model access with Open Router
Aug 31st, 2025
This week brings reasoning controls and simplified AI model access to 120 AI Chat, our native cross-platform AI chat application. We've expanded support for the latest AI models including Gemini Flash 2.5 Image Generation and introduced OpenRouter integration for streamlined multi-model workflows.
Reasoning/thinking control system
120 AI Chat now features reasoning controls with four levels: No Reasoning, Low, Medium, and High. This feature gives users unprecedented control over AI thinking processes within our native macOS interface.
No reasoning mode: Perfect for quick responses and casual conversations where speed takes priority over deep analysis. This mode optimizes for immediate, conversational AI interactions.
Low reasoning: Ideal for standard problem-solving tasks that require some analytical thinking but don't need extensive deliberation. This setting balances response speed with thoughtful analysis.
Medium reasoning: Designed for complex technical discussions, coding challenges, and multi-step problem solving where thorough AI reasoning significantly improves response quality.
High reasoning: Enables maximum AI analytical capability for research tasks, complex creative projects, and sophisticated technical analysis where deep thinking is essential.
System prompt management
We've added system prompt functionality that allows users to customize AI behavior patterns directly within 120 AI Chat's interface.
Workflow customization: Users can define specific AI personalities, response styles, and behavioral guidelines that persist across conversations and model switches. This enables consistent AI interaction patterns tailored to individual workflow needs.
Multi-model compatibility: System prompts work across all integrated AI models, from local Ollama instances to cloud-based services, ensuring consistent AI behavior regardless of underlying model technology.
Gemini Flash 2.5 Image Generation
Google's newest image generation model Nano Banana brings enhanced quality to 120 AI Chat's visual creation capabilities. This latest model offers improved prompt understanding and ideal for rapid prototyping and iterative creative workflows. So far, our team have been really impressed with both quality and speed of this model.
Advanced image workflows: Gemini Flash 2.5 integrates with our multi-threaded conversation system, enabling users to compare image generation approaches across different providers while maintaining conversation context.
Grok Code Fast 1
xAI's specialized coding model Grok Code Fast 1 provides rapid code generation and analysis within 120 AI Chat conversations. This model excels at quick coding tasks, debugging assistance, and technical explanations, complementing our existing AI model ecosystem.
Unified AI access through OpenRouter
We've integrated OpenRouter.io to provide unified access to multiple AI providers through a single API key, simplifying the setup process for 120 AI Chat users who want access to diverse AI capabilities.
One-key access: Instead of managing separate API keys for different providers, users can configure OpenRouter once and gain access to dozens of AI models including the latest releases from OpenAI, Anthropic, Google, and emerging providers.
Model discovery: OpenRouter integration enables automatic access to new models as they become available, ensuring 120 AI Chat users always have access to cutting-edge AI capabilities.
Native macOS AI chat leadership
120 AI Chat continues to lead the native macOS AI chat application space by combining comprehensive model support with intuitive native controls. Our reasoning system, system prompt management, and unified API access create a professional-grade AI interaction platform that leverages macOS's performance and integration capabilities.
Cross-platform expansion: While maintaining our macOS-first development approach, these architectural improvements lay the groundwork for bringing 120 AI Chat's comprehensive AI capabilities to Windows and Linux platforms as native applications.
Performance optimization: All new features maintain our commitment to native performance, ensuring that expanded AI capabilities don't compromise the responsive, smooth experience that sets 120 AI Chat apart from web-based alternatives.
Looking forward
Next week, we will continue working on dedicated Settings windows for image generation models, allowing users to configure batch generation quantities, image dimensions, quality parameters, and other advanced options.
These comprehensive AI integration capabilities establish 120 AI Chat as the definitive native AI chat platform for macOS, with cross-platform expansion planned to bring the same professional-grade AI access to Windows and Linux users through our native application approach.