Poe's new API offers a single, OpenAI-compatible interface to access and switch between top-tier text, image, video, and audio models like Claude, Imagen 4, and Veo 3. All through one API key and billing system.
Hi everyone!
A year ago, Poe was my go-to hub for trying the latest models. While I haven't used it as much lately, it's great to see they're still incredibly fast at supporting the newest models across all modalities—LLMs, voice, and multimodal.
With their new API, some might ask: why not just subscribe to the models directly? I think the answer is similar to why you don't always self-host an open-source model. When Grok first launched, I tried building my own bot with it on Poe, and the official Poe-integrated version was far more stable and robust.
Poe is essentially offering "Poe as a Service"—an optimized, reliable layer on top of the raw models. For any app that needs to switch between different models on the fly, I can't think of many better platforms.
No way, Quora launching its own fast AI chat? That’s wild! I always get lost in threads—having instant, smart answers is a game changer. How quick is it on mobile?
Replies
Awesome work, I used to use Poe a lot, great to see the API get released.
BestPage.ai
No way, Quora launching its own fast AI chat? That’s wild! I always get lost in threads—having instant, smart answers is a game changer. How quick is it on mobile?