Microsoft AI Brings MAI-Voice-2.1 Series to OpenRouter with 23-Language Voice Support

Microsoft AI has launched its MAI-Voice-2.1 and MAI-Voice-2.1-Flash speech models on OpenRouter, a platform for accessing multiple AI models.

Both models can generate speech in 23 languages with the same voice while maintaining native accents. The Flash variant is optimized for low-latency and scale, generating 45 seconds of audio in 150ms, and is priced at $15 per million characters. The standard MAI-Voice-2.1 is priced at $22 per million characters. The models target voice agents, assistants, call centers and other real-time voice and audio scenarios.