askimo

Askimo is a native desktop app that puts several AI providers behind one chat client. The README lists OpenAI, Claude, Gemini, Grok, Ollama, LM Studio, Docker AI, OpenRouter, NVIDIA NIM, Together AI, vLLM and any OpenAI-compatible endpoint, which you can switch between per session, and it supports vision for multimodal models. It describes itself as a real desktop app rather than a web wrapper, with 50 to 300 MB of memory use and long conversations that stay responsive.

Beyond chat, the feature list covers local RAG over folders, files and web URLs with hybrid BM25 and vector retrieval and a classifier that skips retrieval when a query does not need it, web search through DuckDuckGo, Brave, Tavily, Serply or your own SearxNG, Plans that chain multi-step AI pipelines (defined in YAML or generated from a description, exportable as PDF or Word), a script runner for Python, Bash and JavaScript, and MCP tool integration over stdio or HTTP. It can also delegate a goal to Claude Code, OpenAI Codex or Google Antigravity with live streamed tool calls, and lets you define reusable skills once and place them in each agent's skill folder. Dictation and read-aloud work with OpenAI or local Whisper-compatible and Piper endpoints.