I won the Dinoki giveaway! ❤️

Dinoki supports both models using ollama and OpenRouter. Test it out and let us know your thoughts!

🤯 OpenAI 's gpt-oss works incredibly fast on my local machine. I'm not even connected to the internet and is able to produce answers in seconds 👏 well done Sam Altman !

In love with the new gpt-oss! The tool calling abilities are superb 👏 Thank you Sam Altman! 🥹

happy place

Made with Grok 🍅

We had a small hiccup with the release of v1.3.34, the release notes got messed up. We added the ability to toggle all tools for convenience.

Thanks for the mention!

🦕 Introducing Osaurus — an open-source AI server built natively for Apple Silicon (MIT license). ⚡ ~20% faster than Ollama (Swift + Metal, no Electron bloat) 🖥 7MB app with minimal GUI (fetch Hugging Face models, CPU/mem stats) 🔌 OpenAI API compatible → works with Dinoki,

Osaurus – 仅 7MB 的 Apple Silicon 原生开源 LLM 服务，号称比 Ollama 快 20%，基于 Apple 的 MLX 构建，可在 M 系列芯片上实现极致性能。试试~ github.com/dinoki-ai/osau…

通过 Ollama 在 Mac 上使用本地大语言模型，总感觉没有完全发挥出苹果芯片的真正实力。 最近在 GitHub 上看到 Osaurus 这个专为 Apple Silicon 深度优化的本地 LLM 服务器，号称比 Ollama 更快。 基于苹果原生 MLX 框架构建，不仅在流式输出速度上比 Ollama

local LLM server for Apple Silicon with OpenAI API endpoints

Osaurus — an open-source, Apple Silicon–native LLM server (~7 MB) that runs local models on M1/M2 Macs as an Ollama alternative. Privacy-first, OpenAI-compatible, and fast. Try GitHub: github.com/dinoki-ai/osau… What would you run it on — apps, research, or to keep data local?

🙏