Playback speed

Control speaking speed in Rime synthesis using speedAlpha and inlineSpeedAlpha.

Rime allows you to customize the speaking speed of your voice agents, enabling you to:

• Enhance accessibility for diverse audiences.
• Tailor speech to specific applications (for example, slower pace for explanatory materials).
• Optimize the conversation to suit various contexts.

Learn how to adjust the speaking speed using Rime’s APIs.

Adjusting the overall speed

How you adjust overall speaking speed depends on the model.

Coda, Arcana, and Mist v3

Use the timeScaleFactor parameter. A value above 1.0 slows the audio; a value below 1.0 speeds it up.

{
    ...otherParams,
    "text": "Hello, world!",
    "modelId": "coda",
    "timeScaleFactor": 0.85
}

Coda, Arcana, and Mist v3 also accept speedAlpha, but it works in the opposite direction from timeScaleFactor (higher than 1.0 is faster, lower is slower). Going forward we recommend timeScaleFactor for these models — it’s the parameter we’ll continue to support across future flagship models.

timeScaleFactor works over the HTTP API only. On the WebSocket endpoints (/ws, /ws2, /ws3), the parameter is silently ignored — use speedAlpha instead.

Mist v2

Use the speedAlpha parameter. Lower than 1.0 is faster; higher than 1.0 is slower.

{
    ...otherParams,
    "text": "Hello, world!",
    "modelId": "mistv2",
    "speedAlpha": 0.85
}

Adjusting the speed of individual words

Per-word speed adjustment via inlineSpeedAlpha is a Mist-family feature. Coda and Arcana do not support it.

To adjust the speed of individual words or phrases on the Mist family, use the inlineSpeedAlpha parameter. It takes a comma-separated list of speed values applied to words in square brackets.

On mistv2 and mistv3, values < 1.0 speed up speech and values > 1.0 slow it down.

On Mist v3, inlineSpeedAlpha and speedAlpha go in opposite directions — speedAlpha follows the modern flagship direction (higher = faster), but inlineSpeedAlpha follows the legacy Mist direction (lower = faster). Don’t infer one from the other.

For example: “This sentence is [really] [fast]” with inlineSpeedAlpha “0.5, 3” will make “really” fast and “fast” slow on the Mist family.

{
    ...otherParams,
    "modelId": "mistv2",
    "text": "This sentence is [really] [fast]",
    "inlineSpeedAlpha": "0.5, 3"
}

Recommendations

By effectively utilizing Rime’s speed control features, you can create more varied conversational AI experiences.

• Use speedAlpha to adjust the overall speed of the speech.
• Use inlineSpeedAlpha (Mist family only) to adjust the speed of individual words or phrases.
• Experiment with different values to find the best fit for your use case.
• Be mindful that large speed adjustments may impact the naturalness and clarity of the synthesized speech.