Pika SFX模型实现秒级文本转音效

Today we’re putting a spotlight on the innovation behind our Pika SFX model, which delivers producti...

精选理由

Pika Labs推出了Pika SFX模型,能快速将文本转为音效,比其他平台快很多。

AI 摘要

Pika SFX模型处理50个提示时平均耗时0.847秒,完成454秒音频请求测试。对比同类,其速度比ElevenLabs v2快约2.63秒,比Stable Audio 3 SFX快约2.76秒。

原文 · Pika Labs

Today we’re putting a spotlight on the innovation behind our Pika SFX model, which delivers producti...

Today we’re putting a spotlight on the innovation behind our Pika SFX model, which delivers production-ready text-to-SFX in under a second. The stats: across 50 prompts covering 454 seconds of requested audio, Pika SFX completed its locally hosted end-to-end path in 0.847s on average. For context, the fastest hosted APIs in the same benchmark averaged 2.47s (ElevenLabs v2) and 2.64s (Stable Audio 3 SFX). 💬 2 🔄 0 ❤️ 4 👀 1067 📊 2 ⚡