Model details
Flagship Model · 2026.9
Tempolorv4.8
Flagship music-generation model with high-quality 48kHz stereo and tracks up to 360 seconds. Better-sounding, more controllable, faster — taking AI music from "usable" to "genuinely good".
48kHz StereoHierarchical CodecUp to 360sSong CoverPrecise Remix
Try NowAPI Access
Model evolution
v1.0
2024.5
Language Model · 48kHz
The start of controllable generation
v2.0
2024.9
MAR · 44.1kHz
Stronger song-level structure modeling
v3.5
2025.1
Diffusion · 44.1kHz
Refined vocals, more natural structure
v4.5
2026.3
Hierarchical Codec · 48kHz
Song Cover and precise Remix
v5.0
2026.x
Stay tuned
Featured works
View more
36氪waves之夜合作曲官方主题曲
Tempolor v4.6FunkMV
十五运会 AI 主题曲官方主题曲
Tempolor v3.5CinematicAnthem
旋转的洗衣机数字人MV
Tempolor v3.5ElectronicMV
熊出没宣传片
Tempolor v3.5CinematicTrailer
36氪 WISE AI 主题曲大会主题曲
Tempolor v3.5CinematicAnthem
All models
Tempolor v3.5
Diffusion · 2025.1
Diffusion architecture, 44.1kHz stereo, excellent vocal quality, more natural music structure, inference real-time factor below 0.1
Tempolor v2.0
MAR · 2024.9
Masked-autoregressive architecture, 44.1kHz stereo, strengthened song-level structural modeling
Tempolor v1.0
Language Model · 2024.5
Audio language-model architecture, 48kHz stereo, the starting point of Tempolor's controllable music-generation direction