Compare the best AI models for image generation, video generation, coding, chat, and agents. Live rankings from LMSYS Arena with pricing, speed, and performance data.
Live site: squidlab.squidbot.app
SquidLab is a free, fast, mobile-friendly AI model comparison tool. It ranks AI models across five categories:
- Agent Arena — Real-world AI agent performance (task completion, net improvement)
- Chat Arena — Conversational AI ranked by human preference (LMSYS Arena)
- Code Arena — Coding and developer benchmarks
- Image Arena — Best AI models for image generation, ranked by human votes
- Video Arena — Best AI models for video generation, ranked by human votes
Each model shows pricing, context window, speed, and open-source status. Filter by open source, search by name or provider.
Most AI model leaderboards are dense, technical, and hard to read on mobile. SquidLab is built for normal people — especially business owners who want to know "what's the best AI model for X?" without reading a research paper.
- LMSYS Arena (lmarena.ai) — Human preference votes for chat, image, and video
- Agent Arena — Real-world agent task performance
- Provider APIs — Pricing, context windows, speed data from official docs
Data auto-updates daily at 3 AM PT.
80+ models from 24 providers including OpenAI (GPT, Sora), Anthropic (Claude), Google (Gemini, Veo), Meta (Llama), Alibaba (Qwen), Z.ai (GLM), DeepSeek, xAI (Grok), Bytedance, Moonshot (Kimi), Xiaomi, Microsoft, Nvidia, and more.
SquidCircle — makers of SquidBot, an AI system that runs entire business functions and automatically selects the best AI model for each task.
- React + Vite
- Tailwind CSS
- Pure client-side — no backend, no database
MIT