Explore AI model APIs by category

Audio & Speech model APIs

Speech recognition, text-to-speech and audio models for transcription, voice and real-time experiences.

10 available models

Compare model API pricing and capabilities

Open a model page for current prices, endpoints, context limits and related APIs.

All model pricing

Xiaomi

mimo-v2.5-tts-voicedesign

无需参考音频，一句话描述即可生成全新音色。多维度语义理解，覆盖口音、年龄、气质与录音质感。

Context —

Token-based

Suppliers 1

Free

Xiaomi

mimo-v2.5-tts-voiceclone

数秒参考音频，即可高保真复刻目标音色。无需训练或标注，克隆后仍可叠加风格指令与情绪标签。

Context —

Token-based

Suppliers 1

Free

Xiaomi

mimo-v2.5-tts

精品音色语音合成，风格指令精准遵循。内联情绪标签逐字控制，零提示词捕捉角色气质。

Context —

Token-based

Suppliers 1

Free

Xiaomi

mimo-v2.5-asr

中英双语 + 多方言语音识别，无需语言标签，自动识别语码混用中的各语言内容。支持歌词转写，适应强噪声等复杂声学场景。

Context —

Per Request

Suppliers 1

Request$0.011/request

SUNO

suno_wav

Suno 将已有音乐导出为 WAV 音频。

Context —

Token-based

Suppliers 1

Input$75/1MOutput$75/1M

SUNO

suno_stem

Suno 音轨分离接口，用于提取人声或乐器等分轨。

Context —

Token-based

Suppliers 1

Input$75/1MOutput$75/1M

SUNO

suno_music

Suno 音乐生成接口，可根据文字提示创作音乐。

Context —

Per Request

Suppliers 1

Request$0.072/request

Alibaba

qwen3-omni-flash-realtime

千问实时音视频对话模型，可理解文本、图像和音视频输入，并流式输出文本与语音。

Context —

Token-based

Suppliers 2-30%

Input$3.584$2.5088/1MOutput$7.168$5.0176/1M

Alibaba

qwen3-tts-instruct-flash-realtime

千问支持指令控制的实时文本转语音模型，可通过文字指令调整语音风格。

Context —

Token-based

Suppliers 2-30%

Input$75$52.5/1MOutput$75$52.5/1M

Alibaba

qwen3.5-omni-flash-realtime

千问实时音视频对话模型，可理解文本、图像和音视频输入，并流式输出文本与语音。

Context —

Token-based

Suppliers 2-30%

Input$75$52.5/1MOutput$75$52.5/1M