sonilo 0.11.2 → 0.13.0

package/dist/index.d.cts

@@ -100,8 +100,9 @@ interface VideoToMusicParams {
      * finalize-time transcodes and require async. Defaults to m4a. */
     outputFormat?: "m4a" | "wav" | "mp3";
     /** Duck the generated music under the source voice at finalize time.
-     * Default-ON server-side in async mode: leave unset to keep it on, pass
-     * `false` to opt out. Free, best-effort; only valid on `submit()`. */
+     * Default-OFF server-side: leave unset to skip it, pass `true` to run it,
+     * which adds a `ducked` track alongside the unchanged clean `audio`. Free,
+     * best-effort; only valid on `submit()`. */
     ducking?: boolean;
     /** How many distinct music variants to generate in one request (1-10,
      * default 1). Cost scales linearly, and values above 1 are never covered
@@ -109,6 +110,14 @@ interface VideoToMusicParams {
      * by `submit()`) — only meaningful via `submit()`; `stream()`/`generate()`
      * never send it, since they always request a plain stream. */
     variantsNum?: number;
+    /** How strongly the generated music follows the prompt (0-1, API default
+     * 0.5). Lower values let the video lead; higher values follow the prompt
+     * more literally. Free of charge. A generation-time knob, not a
+     * finalize-time one, so it works on the plain stream as well as async —
+     * no `mode` gating. Omitted from the wire when unset so the server
+     * default applies; `0` is a meaningful value and is sent. Out-of-range
+     * values are rejected server-side with a 422. */
+    promptInfluence?: number;
 }
 /** One service's free-trial allowance. `remaining` is already floored at 0,
  * so it is safe to compare directly. */
@@ -282,10 +291,10 @@ interface VideoResult extends BaseTaskResult {
  * | Request | Audio in the returned video |
  * | --- | --- |
  * | neither set | generated music only — the source's own audio is removed |
- * | `keepOriginalSound: true` | full original sound + music ducked under it |
- * | `keepOriginalSound: true, ducking: false` | full original sound + static music mix |
- * | `preserveSpeech: true` | isolated vocals + music ducked under them |
- * | `preserveSpeech: true, ducking: false` | static vocal-forward mix of vocals + music |
+ * | `keepOriginalSound: true` | full original sound + static music mix |
+ * | `keepOriginalSound: true, ducking: true` | full original sound + music ducked under it |
+ * | `preserveSpeech: true` | static vocal-forward mix of vocals + music |
+ * | `preserveSpeech: true, ducking: true` | isolated vocals + music ducked under them |
  *
  * `keepOriginalSound` supersedes `preserveSpeech` — the voice source is a
  * single choice, and keeping the whole track subsumes keeping only the
@@ -309,13 +318,13 @@ interface VideoToVideoMusicParams {
      * source's own audio is removed. Supersedes `preserveSpeech`. */
     keepOriginalSound?: boolean;
     /** How the voice and the generated music are combined — not whether a
-     * voice is kept. Default-ON server-side: leave unset for the dynamic duck
-     * (music dips only while the voice is present), pass `false` for a static
-     * voice-forward mix at a fixed offset. Has no effect when there is no
-     * voice source, i.e. neither `keepOriginalSound` nor `preserveSpeech` is
-     * set. Free and best-effort; silently falls back to generated-audio-only
-     * if the source has no usable audio track, voice isolation fails, or the
-     * mix fails. */
+     * voice is kept. Default-OFF server-side: leave unset for a static
+     * voice-forward mix at a fixed offset, pass `true` for the dynamic duck
+     * (music dips only while the voice is present). Has no effect when there
+     * is no voice source, i.e. neither `keepOriginalSound` nor `preserveSpeech`
+     * is set. Free and best-effort; silently falls back to
+     * generated-audio-only if the source has no usable audio track, voice
+     * isolation fails, or the mix fails. */
     ducking?: boolean;
     /** Keep only the source's isolated speech (not the whole track) in the
      * output. Both this and the legacy `isolateVocals` are accepted and OR'd
@@ -328,6 +337,12 @@ interface VideoToVideoMusicParams {
      * by the free trial. This endpoint is always async, so no extra `mode`
      * gating applies. The result's `videos[]` gets one entry per variant. */
     variantsNum?: number;
+    /** How strongly the generated music follows the prompt (0-1, API default
+     * 0.5). Lower values let the video lead; higher values follow the prompt
+     * more literally. Free of charge. Omitted from the wire when unset so the
+     * server default applies; `0` is a meaningful value and is sent.
+     * Out-of-range values are rejected server-side with a 422. */
+    promptInfluence?: number;
 }
 interface VideoToVideoSfxParams {
     video?: VideoInput;
@@ -367,10 +382,16 @@ interface VideoToVideoSoundParams {
     /** Keep only the source's isolated speech (not the whole track) in the
      * result. Superseded by `keepOriginalSound`. */
     preserveSpeech?: boolean;
-    /** How the voice and the generated bed are combined — not whether a voice
-     * is kept. Default-ON server-side: leave unset for the dynamic duck, pass
-     * `false` for a static voice-forward mix. Has no effect when there is no
-     * voice source. */
+    /** On `videoToVideoSound`, how the voice and the generated bed are
+     * combined — not whether a voice is kept: default-OFF server-side, so an
+     * unset value gives a static voice-forward mix and `true` gives the dynamic
+     * duck, and either way it has no effect without a voice source.
+     *
+     * On `videoToSound` it decides BOTH, because that endpoint has no
+     * `keepOriginalSound`: unset means the source's own speech is left out of
+     * the deliverable entirely, and `true` is what pulls the original track in
+     * and ducks the bed under it. Pass `preserveSpeech` instead to bring in the
+     * isolated vocals rather than the whole track. */
     ducking?: boolean;
     /** How many distinct variants to generate in one request (1-10, default
      * 1). Cost scales linearly, and values above 1 are never covered by the
@@ -389,9 +410,9 @@ interface VideoToSoundParams extends VideoToVideoSoundParams {
     outputFormat?: "wav" | "m4a" | "mp3";
     /** Not available on this endpoint. `keepOriginalSound` only means something
      * when the deliverable is a video whose own audio could be preserved;
-     * `videoToSound` returns generated audio, and its default already keeps the
-     * source voice in the mix. Typed `never` so passing it is a compile error
-     * rather than a field the server silently drops — the mirror of how
+     * `videoToSound` returns generated audio, and reaches the original track
+     * through `ducking: true` instead. Typed `never` so passing it is a compile
+     * error rather than a field the server silently drops — the mirror of how
      * `outputFormat` is kept off `videoToVideoSound`. */
     keepOriginalSound?: never;
 }
@@ -449,8 +470,8 @@ interface DubbingParams {
     languages?: DubbingLanguage[];
     /**
      * Duck the background music/effects bed under the dubbed voice. Default
-     * OFF server-side (the opposite of video-to-music's `ducking`): the bed
-     * is always kept, at a constant level unless this is `true`. Free.
+     * OFF server-side, like every other endpoint's `ducking`: the bed is
+     * always kept, at a constant level unless this is `true`. Free.
      */
     ducking?: boolean;
 }
@@ -703,6 +724,6 @@ declare class RequestTimeoutError extends SoniloError {
 declare function download(media: SfxMedia | string | undefined, fetchFn?: typeof globalThis.fetch, timeout?: number): Promise<Uint8Array>;
 
 /** The SDK's own version. Generated by scripts/sync-versions.mjs — do not edit. */
-declare const VERSION = "0.11.2";
+declare const VERSION = "0.13.0";
 
 export { APIError, type AccountServices, type AudioChunkEvent, AuthenticationError, BadRequestError, type BaseTaskResult, type CompleteEvent, type CostEvent, type CostInfo, DEFAULT_TIMEOUT_MS, type DailyUsage, type DubbingLanguage, type DubbingParams, type DubbingResult, type ErrorEvent, GenerationError, type MusicMediaEntry, type MusicMuxEntry, type MusicTaskResult, type MusicTitle, PaymentRequiredError, RateLimitError, RequestTimeoutError, type Segment, type SegmentLabel, type SfxAudioFormat, type SfxError, type SfxMedia, type SfxResult, type SfxSegment, type SfxTask, SoniloClient, type SoniloClientOptions, SoniloError, type SoundOutputEntry, type SoundResult, type StreamEvent, TaskFailedError, TaskTimeoutError, type TextToMusicParams, type TextToSfxParams, type TitleEvent, type Track, TrialExhaustedError, type TrialQuota, type UnknownEvent, type UsageResponse, type UsageSummary, VERSION, type VideoInput, type VideoResult, type VideoToMusicParams, type VideoToSfxParams, type VideoToSoundParams, type VideoToVideoMusicParams, type VideoToVideoSfxParams, type VideoToVideoSoundParams, type WaitOptions, download, isAudioChunkEvent, isErrorEvent };

package/dist/index.d.ts

@@ -100,8 +100,9 @@ interface VideoToMusicParams {
      * finalize-time transcodes and require async. Defaults to m4a. */
     outputFormat?: "m4a" | "wav" | "mp3";
     /** Duck the generated music under the source voice at finalize time.
-     * Default-ON server-side in async mode: leave unset to keep it on, pass
-     * `false` to opt out. Free, best-effort; only valid on `submit()`. */
+     * Default-OFF server-side: leave unset to skip it, pass `true` to run it,
+     * which adds a `ducked` track alongside the unchanged clean `audio`. Free,
+     * best-effort; only valid on `submit()`. */
     ducking?: boolean;
     /** How many distinct music variants to generate in one request (1-10,
      * default 1). Cost scales linearly, and values above 1 are never covered
@@ -109,6 +110,14 @@ interface VideoToMusicParams {
      * by `submit()`) — only meaningful via `submit()`; `stream()`/`generate()`
      * never send it, since they always request a plain stream. */
     variantsNum?: number;
+    /** How strongly the generated music follows the prompt (0-1, API default
+     * 0.5). Lower values let the video lead; higher values follow the prompt
+     * more literally. Free of charge. A generation-time knob, not a
+     * finalize-time one, so it works on the plain stream as well as async —
+     * no `mode` gating. Omitted from the wire when unset so the server
+     * default applies; `0` is a meaningful value and is sent. Out-of-range
+     * values are rejected server-side with a 422. */
+    promptInfluence?: number;
 }
 /** One service's free-trial allowance. `remaining` is already floored at 0,
  * so it is safe to compare directly. */
@@ -282,10 +291,10 @@ interface VideoResult extends BaseTaskResult {
  * | Request | Audio in the returned video |
  * | --- | --- |
  * | neither set | generated music only — the source's own audio is removed |
- * | `keepOriginalSound: true` | full original sound + music ducked under it |
- * | `keepOriginalSound: true, ducking: false` | full original sound + static music mix |
- * | `preserveSpeech: true` | isolated vocals + music ducked under them |
- * | `preserveSpeech: true, ducking: false` | static vocal-forward mix of vocals + music |
+ * | `keepOriginalSound: true` | full original sound + static music mix |
+ * | `keepOriginalSound: true, ducking: true` | full original sound + music ducked under it |
+ * | `preserveSpeech: true` | static vocal-forward mix of vocals + music |
+ * | `preserveSpeech: true, ducking: true` | isolated vocals + music ducked under them |
  *
  * `keepOriginalSound` supersedes `preserveSpeech` — the voice source is a
  * single choice, and keeping the whole track subsumes keeping only the
@@ -309,13 +318,13 @@ interface VideoToVideoMusicParams {
      * source's own audio is removed. Supersedes `preserveSpeech`. */
     keepOriginalSound?: boolean;
     /** How the voice and the generated music are combined — not whether a
-     * voice is kept. Default-ON server-side: leave unset for the dynamic duck
-     * (music dips only while the voice is present), pass `false` for a static
-     * voice-forward mix at a fixed offset. Has no effect when there is no
-     * voice source, i.e. neither `keepOriginalSound` nor `preserveSpeech` is
-     * set. Free and best-effort; silently falls back to generated-audio-only
-     * if the source has no usable audio track, voice isolation fails, or the
-     * mix fails. */
+     * voice is kept. Default-OFF server-side: leave unset for a static
+     * voice-forward mix at a fixed offset, pass `true` for the dynamic duck
+     * (music dips only while the voice is present). Has no effect when there
+     * is no voice source, i.e. neither `keepOriginalSound` nor `preserveSpeech`
+     * is set. Free and best-effort; silently falls back to
+     * generated-audio-only if the source has no usable audio track, voice
+     * isolation fails, or the mix fails. */
     ducking?: boolean;
     /** Keep only the source's isolated speech (not the whole track) in the
      * output. Both this and the legacy `isolateVocals` are accepted and OR'd
@@ -328,6 +337,12 @@ interface VideoToVideoMusicParams {
      * by the free trial. This endpoint is always async, so no extra `mode`
      * gating applies. The result's `videos[]` gets one entry per variant. */
     variantsNum?: number;
+    /** How strongly the generated music follows the prompt (0-1, API default
+     * 0.5). Lower values let the video lead; higher values follow the prompt
+     * more literally. Free of charge. Omitted from the wire when unset so the
+     * server default applies; `0` is a meaningful value and is sent.
+     * Out-of-range values are rejected server-side with a 422. */
+    promptInfluence?: number;
 }
 interface VideoToVideoSfxParams {
     video?: VideoInput;
@@ -367,10 +382,16 @@ interface VideoToVideoSoundParams {
     /** Keep only the source's isolated speech (not the whole track) in the
      * result. Superseded by `keepOriginalSound`. */
     preserveSpeech?: boolean;
-    /** How the voice and the generated bed are combined — not whether a voice
-     * is kept. Default-ON server-side: leave unset for the dynamic duck, pass
-     * `false` for a static voice-forward mix. Has no effect when there is no
-     * voice source. */
+    /** On `videoToVideoSound`, how the voice and the generated bed are
+     * combined — not whether a voice is kept: default-OFF server-side, so an
+     * unset value gives a static voice-forward mix and `true` gives the dynamic
+     * duck, and either way it has no effect without a voice source.
+     *
+     * On `videoToSound` it decides BOTH, because that endpoint has no
+     * `keepOriginalSound`: unset means the source's own speech is left out of
+     * the deliverable entirely, and `true` is what pulls the original track in
+     * and ducks the bed under it. Pass `preserveSpeech` instead to bring in the
+     * isolated vocals rather than the whole track. */
     ducking?: boolean;
     /** How many distinct variants to generate in one request (1-10, default
      * 1). Cost scales linearly, and values above 1 are never covered by the
@@ -389,9 +410,9 @@ interface VideoToSoundParams extends VideoToVideoSoundParams {
     outputFormat?: "wav" | "m4a" | "mp3";
     /** Not available on this endpoint. `keepOriginalSound` only means something
      * when the deliverable is a video whose own audio could be preserved;
-     * `videoToSound` returns generated audio, and its default already keeps the
-     * source voice in the mix. Typed `never` so passing it is a compile error
-     * rather than a field the server silently drops — the mirror of how
+     * `videoToSound` returns generated audio, and reaches the original track
+     * through `ducking: true` instead. Typed `never` so passing it is a compile
+     * error rather than a field the server silently drops — the mirror of how
      * `outputFormat` is kept off `videoToVideoSound`. */
     keepOriginalSound?: never;
 }
@@ -449,8 +470,8 @@ interface DubbingParams {
     languages?: DubbingLanguage[];
     /**
      * Duck the background music/effects bed under the dubbed voice. Default
-     * OFF server-side (the opposite of video-to-music's `ducking`): the bed
-     * is always kept, at a constant level unless this is `true`. Free.
+     * OFF server-side, like every other endpoint's `ducking`: the bed is
+     * always kept, at a constant level unless this is `true`. Free.
      */
     ducking?: boolean;
 }
@@ -703,6 +724,6 @@ declare class RequestTimeoutError extends SoniloError {
 declare function download(media: SfxMedia | string | undefined, fetchFn?: typeof globalThis.fetch, timeout?: number): Promise<Uint8Array>;
 
 /** The SDK's own version. Generated by scripts/sync-versions.mjs — do not edit. */
-declare const VERSION = "0.11.2";
+declare const VERSION = "0.13.0";
 
 export { APIError, type AccountServices, type AudioChunkEvent, AuthenticationError, BadRequestError, type BaseTaskResult, type CompleteEvent, type CostEvent, type CostInfo, DEFAULT_TIMEOUT_MS, type DailyUsage, type DubbingLanguage, type DubbingParams, type DubbingResult, type ErrorEvent, GenerationError, type MusicMediaEntry, type MusicMuxEntry, type MusicTaskResult, type MusicTitle, PaymentRequiredError, RateLimitError, RequestTimeoutError, type Segment, type SegmentLabel, type SfxAudioFormat, type SfxError, type SfxMedia, type SfxResult, type SfxSegment, type SfxTask, SoniloClient, type SoniloClientOptions, SoniloError, type SoundOutputEntry, type SoundResult, type StreamEvent, TaskFailedError, TaskTimeoutError, type TextToMusicParams, type TextToSfxParams, type TitleEvent, type Track, TrialExhaustedError, type TrialQuota, type UnknownEvent, type UsageResponse, type UsageSummary, VERSION, type VideoInput, type VideoResult, type VideoToMusicParams, type VideoToSfxParams, type VideoToSoundParams, type VideoToVideoMusicParams, type VideoToVideoSfxParams, type VideoToVideoSoundParams, type WaitOptions, download, isAudioChunkEvent, isErrorEvent };

package/dist/index.js

@@ -392,6 +392,9 @@ var VideoToMusic = class {
     if (params.segments !== void 0) {
       form.set("segments", JSON.stringify(params.segments));
     }
+    if (params.promptInfluence !== void 0) {
+      form.set("prompt_influence", String(params.promptInfluence));
+    }
     const res = await this.client.request(
       "/v1/video-to-music",
       { method: "POST", body: form, signal: params.signal },
@@ -452,6 +455,9 @@ var VideoToMusic = class {
     if (params.variantsNum !== void 0) {
       form.set("variants_num", String(params.variantsNum));
     }
+    if (params.promptInfluence !== void 0) {
+      form.set("prompt_influence", String(params.promptInfluence));
+    }
     const res = await this.client.request("/v1/video-to-music", {
       method: "POST",
       body: form
@@ -548,6 +554,9 @@ var VideoToVideoMusic = class {
     if (params.variantsNum !== void 0) {
       form.set("variants_num", String(params.variantsNum));
     }
+    if (params.promptInfluence !== void 0) {
+      form.set("prompt_influence", String(params.promptInfluence));
+    }
     const res = await this.client.request("/v1/video-to-video-music", {
       method: "POST",
       body: form
@@ -705,7 +714,7 @@ var Dubbing = class {
 };
 
 // src/version.ts
-var VERSION = "0.11.2";
+var VERSION = "0.13.0";
 
 // src/client.ts
 var DEFAULT_BASE_URL = "https://api.sonilo.com";