https://www.youngju.dev/blog 천천히 올바르게. AI Researcher & DevOps Engineer Youngju's tech blog. GPU/CUDA, LLM, MLOps, Kubernetes AI workloads, distributed training, and data engineering. ko fjvbn2003@gmail.com (Youngju Kim) fjvbn2003@gmail.com (Youngju Kim) Fri, 15 May 2026 00:00:00 GMT https://www.youngju.dev/blog/culture/2026-05-15-voice-ai-2026-elevenlabs-cartesia-sesame-whisper-deepgram-parakeet-deep-dive.en https://www.youngju.dev/blog/culture/2026-05-15-voice-ai-2026-elevenlabs-cartesia-sesame-whisper-deepgram-parakeet-deep-dive.en In October 2024 Whisper Large v3 Turbo got 8x faster, Cartesia (built by the Mamba authors) hit sub-90ms TTS, and Sesame from Brendan Iribe launched "voice presence." By May 2026 the TTS, STT, and realtime-agent axes have all exploded at once. From ElevenLabs V3 to NVIDIA Parakeet 1.1, VOICEVOX, F5-TTS, and Vapi/Retell — who does what well, and how to choose. Fri, 15 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) voice-aittssttasrelevenlabscartesiasesamewhisperdeepgramparakeetassemblyaivoicevoxvapiretell2026deep-diveenglish https://www.youngju.dev/blog/culture/2026-05-15-voice-ai-2026-elevenlabs-cartesia-sesame-whisper-deepgram-parakeet-deep-dive.ja https://www.youngju.dev/blog/culture/2026-05-15-voice-ai-2026-elevenlabs-cartesia-sesame-whisper-deepgram-parakeet-deep-dive.ja 2024年10月にWhisper Large v3 Turboが8倍速くなり、Mamba著者たちが立ち上げたCartesiaが90ms未満TTSを出し、Sesame(Brendan Iribe)が「voice presence」で殴り込み、2026年5月時点で音声AIのTTS・STT・リアルタイムエージェントの3軸が同時に爆発した。ElevenLabs V3からNVIDIA Parakeet 1.1、VOICEVOX、F5-TTS、Vapi/Retellまで、誰が何に強く、どう選ぶかを整理する。 Fri, 15 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) voice-aittssttasrelevenlabscartesiasesamewhisperdeepgramparakeetassemblyaivoicevoxvapiretell2026deep-dive日本語 https://www.youngju.dev/blog/culture/2026-05-15-voice-ai-2026-elevenlabs-cartesia-sesame-whisper-deepgram-parakeet-deep-dive https://www.youngju.dev/blog/culture/2026-05-15-voice-ai-2026-elevenlabs-cartesia-sesame-whisper-deepgram-parakeet-deep-dive 2024년 10월 Whisper Large v3 Turbo가 8배 빨라지고, Cartesia가 Mamba 저자들 손으로 90ms TTS를 만들고, Sesame의 Brendan Iribe가 "voice presence"를 들고 나오면서 2026년 음성 AI는 TTS·STT·실시간 에이전트 세 축이 모두 폭발했다. ElevenLabs V3부터 NVIDIA Parakeet 1.1, VOICEVOX, F5-TTS, Vapi/Retell까지 — 누가 무엇을 잘하고 무엇을 골라야 하는지 정리한다. Fri, 15 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) voice-aittssttasrelevenlabscartesiasesamewhisperdeepgramparakeetassemblyaivoicevoxvapiretell2026deep-dive