Mistral: Voxtral Small 24B 2507

Mistral: Voxtral Small 24B 2507 is a cutting-edge AI model that builds upon the robust foundation of Mistral Small 3. It integrates state-of-the-art audio input capabilities, making it exceptionally proficient in tasks such as speech transcription, language translation from audio, and comprehensive audio understanding. This model maintains the best-in-class text performance expected from Mistral, offering a versatile solution for both audio and text-based AI applications. Designed for a wide range of uses, Voxtral Small 24B 2507 is ideal for chat applications, code generation, and translation services. It supports a substantial context window of 32K tokens and can generate outputs up to 4K tokens. Input audio is priced at $100 per million seconds, with text pricing at $0.10 per 1M input tokens and $0.30 per 1M output tokens. This model offers 'functions', 'code', and 'streaming' capabilities, providing developers with powerful tools for integration. Access it for free on Multi AI.