Dinoki – Privacy-First Desktop AI with Pixel Pets (6MB Native)
Hey HN! Dinoki is a 6MB native AI assistant with pixel pets that live on your desktop. Built for power users who want privacy, performance – and a bit of fun. See it in action: https://www.youtube.com/watch?v=V_pv2BU0CZA What makes it different: - 6MB native app (SwiftUI for macOS, WPF for Windows) – starts instantly - Zero telemetry – your API keys and data stay local - Pixel art companions with physics-based behaviors - Flexible AI support – OpenAI, Anthropic, OpenRouter (300+ models), or Ollama (offline) Why pixel pets in an AI assistant? Beyond being fun, pixel art is…
In plain words
Dinoki is a 6MB native AI assistant for macOS and Windows that combines a desktop chatbot with animated pixel pet companions. It prioritizes privacy by storing API keys and data locally with zero telemetry, while supporting multiple AI providers including OpenAI, Anthropic, OpenRouter, and offline Ollama. The lightweight application launches instantly and features physics-based pixel art animations designed to reduce isolation during long work sessions.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hey HN! Dinoki is a 6MB native AI assistant with pixel pets that live on your desktop. Built for power users who want privacy, performance – and a bit of fun. See it in action: https://www.youtube.com/watch?v=V_pv2BU0CZA What makes it different: - 6MB native app (SwiftUI for macOS, WPF for Windows) – starts instantly - Zero telemetry – your API keys and data stay local - Pixel art companions with physics-based behaviors - Flexible AI support – OpenAI, Anthropic, OpenRouter (300+ models), or Ollama (offline) Why pixel pets in an AI assistant? Beyond being fun, pixel art is incredibly efficient – minimal CPU usage for animations while adding personality to your workspace. After months of testing, users tell us the companions make long work sessions feel less isolating. They develop behaviors, respond to your activity, and yes, they bounce around thanks to physics-based movement. The serious features: - 20+ productivity tools: web research, Slack automation, screenshot OCR - MCP (Model Context Protocol) support for custom integrations - Agent mode for autonomous task execution - Persistent conversation memory - Instant startup (not Electron!) Real use cases from our users: - Developer running local RAG with Qwen3 via Ollama – a private knowledge base with a pixel companion - Home lab enthusiast who set up an AI server for their family – each member has their own Dinoki instance - Remote engineers using Agent mode to monitor Slack channels while their pixel buddy reflects their mood - Researchers who love that their sensitive data never leaves their machine, yet they get GPT-4-level assistance Pricing: - Free: Full chat, one Dinoki character, bring-your-own provider - Pro ($25 lifetime): All 4 characters, automation tools, multi-instance support We're a two-person team that believes AI tools should be powerful, private, and enjoyable to use. The 6MB size isn't just a flex—it's proof that modern software doesn't need to be bloated. Happy to answer any questions about implementation, how we got the size so small, or why we think personality matters in productivity tools! https://dinoki.ai/ P.S. Something big is coming for retro pixel art fans. Current users will get it free.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, August 2025
the whole month →
- IS
I built the world's most impractical 1000-pixel display and anyone in the world can draw on it. It draws a single pixel at a time and takes 30-60 minutes to complete a single image. Anyone can participate in the project by voting for the next image to be drawn, and submitting images. https://kilopx.com/
Work · 2025 · benholmen.com
- KT
Kitten TTS is an open-source series of tiny and expressive text-to-speech models for on-device applications. We are excited to launch a preview of our smallest model, which is less than 25 MB. This model has 15M parameters. This release supports English text-to-speech applications in eight voices: four male and four female. The model is quantized to int8 + fp16, and it uses onnx for runtime. The model is designed to run literally anywhere eg. raspberry pi, low-end smartphones, wearables, browsers etc. No GPU required! We're releasing this to give early users a sense of the latency and voices…
Dev tools · 2025 · github.com
- IW
I was wondering how I can arrange objects along a spherical helix path, and read some articles on it. I ended up learning about parametric equations again, and make this visualization to document what I learned: https://visualrambling.space/moving-objects-in-3d/ feel free to visit and let me know what you think!
Life & fun · 2025 · visualrambling.space
- TC
For HTML Day 2025 [1], I made a web service that displays the current sky at your approximate location as a CSS gradient. Colours are simulated on-demand using atmospheric absorption and scattering coefficients. Updates every minute, without the use of client-side JavaScript. Source code and additional information is available on GitHub: https://github.com/dnlzro/horizon [1] https://html.energy/html-day/2025/index.html
Dev tools · 2025 · sky.dlazaro.ca