MASIH LANGGANAN

@zeven41

1.5K followers · usually 23.5K views a post

Views

45.3K

Likes

2.4K

Comments

53

Shares

225

SoundTikTok sound · Chill Vibes

Promoted product

VoxCPM2

VoxCPM2: Tokenizer-Free TTS for Multilingual Speech Generation, Creative Voice Design, and True-to-Life Cloning - OpenBMB/VoxCPM

Why it worked

The video effectively targets a pain point (monthly subscription costs for AI voice services) and offers a compelling, free, and open-source solution. It clearly explains the benefits and features of VoxCPM2, making it an attractive alternative for creators looking to save money while maintaining quality.

Summary

The video introduces VoxCPM2, an open-source AI text-to-speech model that offers high-quality voice generation in 30 languages, including Indonesian, with a commercial-use license. It highlights features like voice cloning and design, and suggests it as a free alternative to paid services like ElevenLabs for users with a capable GPU.

Structure

1Introduction to VoxCPM2 as an open-source alternative to paid AI voice services.
2Explanation of VoxCPM2's features: multilingual support, high-quality audio, and commercial license.
3Demonstration of advanced features like voice cloning and design.

Original caption

Bayar AI voice tiap bulan? 🤔 Mungkin saatnya mulai lirik alternatif open-source. VoxCPM2 bisa generate voice berkualitas tinggi, support Bahasa Indonesia, dan bebas dipakai secara komersial. #AI #ArtificialIntelligence #GitHub #OpenSource #ElevenLabs #TextToSpeech #VoiceAI #ContentCreator #TechTok #Programming #fyp

More from @zeven41

Product placement

GitHub: platform for software development, where the VoxCPM2 project is hosted. It merely appears, and removing it would not change what happens in the video. VoxCPM2: an open-source AI text-to-speech model. It ACTS as the subject of the video, as its features and capabilities are being demonstrated and discussed. Removing it would fundamentally change the video. ElevenLabs: a paid AI voice service. It is mentioned as a point of comparison, and removing it would not change what happens in the video.

Call to action

Boleh dikomen

On-screen text

MASIH LANGGANAN ELEVEN LABS BUAT AI SEBELUM PERPANJANG SUBSCRIPTION LIHAT REPO GITHUB INI DULU YA NAMANYA VOXCPM NAMANYA VOXCPM2 INI ADALAH AI TEXT TO SPEECH OPEN SOURCE TENTUNYA SUARA BERKUALITAS 48 KHZ MENDUKUNG 30 BAHASA TERMASUK BAHASA INDONESIA DAN LISENSINYA APACHE 2.0 JADI BOLEH DIPAKAI UNTUK KEBUTUHAN KOMERSIAL NAH YANG LEBIH SUARA BARU HANYA DARI DESKRIPSI TEKS ATAU CLONE REFERENCE CLIP DAN KECEPATANNYA TAPI ADA SATU KALAU KALIAN MAU MENJALANKAN MODEL UTAMANYA DI PC SENDIRI MINIMAL KALIAN BUTUH GPU DENGAN SEKITAR 8 GB VRAM NAH KALAU PC INI BISA JADI ALTERNATIF MENARIK BUAT BIKIN VOICE OVER AI TANPA HARUS TERUS BAYAR LAYANAN BULANAN OKE TEMAN-TEMAN KALAU KALIAN PUNYA REKOMENDASI REPO GITHUB YANG MENARIK BOLEH DIKOMEN NEXT VIDEO AKAN GUA REVIEW THANK

Transcript

Bayar AI voice tiap bulan? 🤔 Mungkin saatnya mulai lirik alternatif open-source. VoxCPM2 bisa generate voice berkualitas tinggi, support Bahasa Indonesia, dan bebas dipakai secara komersial. Lihat repo GitHub ini dulu ya, teman-teman. Namanya VoxCPM, namanya VoxCPM2. Ini adalah AI text to speech open source tentunya, yang bisa menghasilkan suara berkualitas 48 KHz, mendukung 30 bahasa, termasuk Bahasa Indonesia. Dan lisensinya Apache 2.0, jadi boleh dipakai untuk kebutuhan komersial. Nah, yang lebih keren lagi, kalian bisa bikin suara baru hanya dari deskripsi teks atau clone reference clip, dengan optional style guidance to steer emotion, pace, and expression while preserving the original timbre. Clone any voice from a short reference clip, with optional style guidance to steer emotion, pace, and expression while preserving the original timbre. Reproduce every vocal nuance: provide both reference audio and its transcript, and the model continues seamlessly from the reference, faithfully preserving every vocal detail - timbre, rhythm, emotion, and style (same as VoxCPM1.5). Accepts 16kHz reference audio and directly outputs 48kHz studio-quality audio via AudioVAE V2's asymmetric encode/decode design, with built-in super-resolution - no external upsampler needed. Automatically infers appropriate prosody and expressiveness from text content. RTF as low as ~0.3 on NVIDIA RTX 4090, and ~0.13 accelerated by Nano-vLLM or vLLM-Omni - official vLLM omni-modal serving for VoxCPM2 with PagedAttention and an OpenAI-compatible API. Weights and code released under the Apache 2.0 license, free for commercial use. Nah, kalau PC kalian punya GPU dengan sekitar 8 GB VRAM, ini bisa jadi alternatif menarik buat bikin voice over AI tanpa harus terus bayar layanan bulanan. Oke teman-teman, kalau kalian punya rekomendasi repo GitHub yang menarik, boleh dikomen. Next video akan gua review. Thank you.