Qwen introduces the real-time speech recognition model Fun-ASR-Realtime, reducing the first-word latency to the millisecond level and achieving a smooth "speak and feedback immediately" interaction. Its recognition accuracy is close to that of offline models, achieving high precision while breaking through the real-time performance bottleneck, marking a new height in voice interaction experience.
Tesla China releases a promotional video for a new in-car voice assistant, showcasing intelligent upgrades and new features, sparking industry interest. Previously criticized for lacking localized technology, this update signals a major transformation in its voice system in China.....
Smartphone operating systems may face a revolutionary change. At the OpenAI Voice Hack Night, a tech team unveiled a prototype of an 'Agentic operating system,' overturning the traditional app ecosystem. Its core concept is 'UI as system,' where phones lack conventional apps, and all interfaces are generated in real-time by on-device models based on user commands.....
The platform announced that multiple older large language models, including GPT-5.2 and GPT-5.3-Codex, will be forcibly retired on June 2, 2026, with a full transition to the flagship GPT-5.5 model. This move has sparked controversy among developers, who report significant performance degradation in the new model and have publicly voiced complaints and backlash.....
A unified API that can generate images, synthesize voices, transcribe audio and video, and provide low-cost access to open-source models.
An all-in-one AI platform with over 30 models, enabling easy creation of videos, images, and voiceovers.
Voiceley can perform AI voice cloning quickly and for free, and also generate voices using voice models.
Provides ASR, TTS, and LLM models for voice AI, which can be tested and deployed for real-time applications.
Google
$0.49
Input tokens/M
$2.1
Output tokens/M
1k
Context Length
Openai
$2.8
$11.2
Xai
$1.4
$3.5
2k
$7.7
$30.8
200
-
Anthropic
$105
$525
$0.7
$7
$35
$17.5
$21
Alibaba
$2
$20
$4
$16
Baidu
128
$6
$24
256
$1
$10
unsloth
Spark-TTS is an efficient text-to-speech system based on large language models (LLM), supporting bilingual synthesis in Chinese and English with zero-shot voice cloning.
IbrahimAmin
This is an automatic speech recognition model fine-tuned based on the wav2vec2-large-xlsr-53 architecture, specifically optimized for Egyptian Arabic, Modern Standard Arabic, and Gulf/Levantine Arabic. The model is trained on multiple Arabic speech datasets and achieves a word error rate of 27.20% on the Common Voice 17.0 Arabic test set, outperforming many similar models.
maitrix-org
Voila is a large family of speech-language foundation models designed to enhance human-computer interaction, supporting real-time, low-latency voice interaction and multilingual processing.
dangvansam
VietTTS is an open-source toolkit providing powerful Vietnamese TTS models, supporting natural speech synthesis and voice cloning.
TeamSpeak MCP is a service based on the Model Context Protocol for controlling TeamSpeak servers through AI models (such as Claude), providing comprehensive channel management, user permission control, voice adjustment, and other functions.
TeamSpeak MCP is a server control tool based on the Model Context Protocol, specifically designed to allow AI models (such as Claude) to manage TeamSpeak voice servers. It provides 39 functional tools, covering all - around operations such as user management, channel control, and permission configuration. It supports multiple deployment methods (PyPI/Docker/local) to achieve automated TeamSpeak management.
A local audio transcription MCP server based on whisper.cpp, supporting multiple models and audio formats. It can work with the Apple Voice Memo MCP to implement a complete voice workflow.
An intelligent conversational robot project based on large models, supporting multi - platform access and multiple AI models, with text, voice, image processing, and plugin expansion capabilities, and can customize enterprise AI applications.