Dev.to
7/13/2026
The original title is "AutarkChat: A developer workspace for side-by-side LLM comparison and evaluation"
Original: I Got Tired of Comparing AI Models Across Multiple Browser Tabs, So I Built AutarkChat
Short summary
AutarkChat is a developer workspace for comparing, testing, and evaluating multiple LLMs simultaneously. Users send a single prompt to multiple models and stream responses independently, with per-response token counts and cost metrics. It supports custom OpenAI-compatible endpoints (OpenAI, DeepSeek, Mistral, Groq, etc.) and features reusable profiles, global instructions, and resilient streaming that preserves successful responses even when one provider fails.
- •Send one prompt to multiple models and compare streaming responses side by side
- •Tracks token usage, cost, and latency per model for informed trade-off decisions
- •Supports any OpenAI-compatible endpoint; built with Next.js 15, TypeScript, MongoDB
Generated with AI, which can make mistakes.
Is this a good recommendation for you?
