Show HN: Agnes AI – Free multimodal API (text, image, video), OpenAI-compatible
Agnes AI, a Singapore-based lab, launched a free multimodal API (text, image, video) that is OpenAI-compatible. It offers models like Agnes-2.0-Flash with a 512k context window and requires no payment information. The service hit 4 trillion API calls in its first week and ranks among the top 10 AI labs on benchmarks.
Background
- Agnes AI is a Singapore-based lab offering a free, OpenAI-compatible API for text, image, and video generation/editing — a rare all-in-one, no-cost offering in a field where most multimodal APIs charge per token or per image.
- "OpenAI-compatible" means developers can swap a single line of code (the base URL) to use Agnes instead of OpenAI's paid API, making integration trivial for existing projects.
- The text model (Agnes-2.0-Flash) boasts a 512k context window (soon 1M) — far larger than most free or paid models, enabling processing of very long documents or codebases.
- Being free with no payment info required is unusual and likely a strategy to gather real-world usage data, feedback, and community adoption before monetization.
- The post references benchmarks like Claw-Eval and Artificial Analysis, independent third-party rankings that help establish credibility for an otherwise unknown lab.
- The broader context: the AI API market is dominated by paid services (OpenAI, Anthropic, Google) and heavily rate-limited free tiers. A fully free multimodal API, if reliable, could significantly lower the barrier for hobbyists, indie developers, and prototyping.