Stability AI has released a new text-to-music artificial intelligence model called AudioSparx, which now supports its Stable Audio product. Compared to previous state-of-the-art AI music generators, this new model is capable of generating high-fidelity, long-form stereo music with more variations and structures. At the core of AudioSparx 1.0 is a latent diffusion model that can quickly generate music based on text prompts. Unlike previous iterations that could only generate 30-second audio, the new model utilizes an enhanced control system to reliably output stereo music lasting up to 95 seconds at CD-quality 44.1kHz sampling rate. Most importantly, AudioSparx 1.0 can mimic the overall form and progression of complete songs in a way that competitors cannot match. The generated tracks include recognizable intros, verse/chorus patterns, transitions, instrument breaks, and endings. This musicality demonstrates a sophisticated understanding of basic song structures. In addition to music, AudioSparx 1.0 is the first AI system capable of generating true 44.1kHz stereo sound effects based on text prompts. Users can request sounds like "outdoor forest bird chirping" and receive immersive binaural audio. Enhancing the prompts with "high-quality, stereo" yields the best results. AudioSparx 1.0 excels in generating variable-length music and sound, representing an outstanding integration of multiple audio synthesis capabilities into a single model. This unified capability stems from Stability AI's general training program, which does not strictly differentiate between musical and non-musical sources. Overall, the innovative technology employed by AudioSparx 1.0 holds promise as an adaptive tool for professional creators in audio production. The model can provide a wide range of meticulously arranged music and sound that surpasses previous benchmarks and fulfills requirements that were previously achievable only through manual production. It showcases Stability AI's commitment to advancing artificial intelligence to match human capabilities.