iSpeech vs VoxCPM

iSpeech provides text-to-speech and automated speech recognition through APIs, SDKs, and online tools. Its speech API can synthesize audio with multiple voices, formats, bitrates, frequencies, and playback speeds, and supports MathML and SSML. The API can also transcribe submitted audio across multiple languages and recognition models; iSpeech says custom models can be created. The homepage lists support for more than 30 languages. Audio downloads are available in AIFF, MP3, OGG, WMA, FLAC, WAV, A-law, μ-law, VOX, and MP4 formats. Listed access includes Android, iOS, Linux, macOS, web, Windows, API, and extension. The Basic plan is free for personal use, and commercial use is listed as available. Plus costs $2.95 a month, billed monthly, and includes 30-minute recordings, approximately 4,167 words. Premium costs $3.95 a month, billed monthly, with 12-hour recordings, approximately 100,000 words. Pay Per Use costs $50 a month for dynamic TTS usage only and includes 2,000 credits. Pay Per Install costs $0.25 a month and has a 10,000-install minimum.

VoxCPM is an open-source text-to-speech system whose VoxCPM2 model generates expressive multilingual speech. It supports 30 languages and lists nine Chinese dialects. Voice Design can create a voice from a natural-language description without reference audio, and voice cloning is supported. Output is 48kHz audio, reference audio is accepted at 16kHz, and WAV is listed as an export format. A streaming generation API is available; the repository reports a real-time factor as low as about 0.3 on an NVIDIA RTX 4090. Full and LoRA fine-tuning can adapt the model from 5–10 minutes of audio. Documented deployment includes Python and CLI use, a local web demo, production serving and on-device inference. vLLM-Omni exposes an OpenAI-compatible /v1/audio/speech endpoint, while Nano-vLLM supports concurrent requests. Apple Silicon MPS and C++ inference on CPU, Metal, CUDA or Vulkan are documented. Code and weights use Apache-2.0 and are free for commercial use. Python installation requires Python 3.10 or later but below 3.13, PyTorch 2.5.0 or later and CUDA 12.0 or later. Voice Design and cloning results can vary, and unsupported languages may require testing or fine-tuning. VoxCPM is headquartered in Beijing, China.