chore: update OpenAPI spec from monorepo (#943)

Co-authored-by: OpenRouter SDK Bot <sdk-bot@openrouter.ai>

Author: openrouter-docs-sync[bot]

.speakeasy/in.openapi.yaml

@@ -23804,7 +23804,9 @@ components:
           - 'temperature'
           - 'top_p'
           - 'max_tokens'
+        supports_image_reference: false
         supports_implicit_caching: true
+        supports_multiple_audio_references: false
         supports_tool_choice:
           auto: true
           function: true
@@ -24096,8 +24098,16 @@ components:
           items:
             $ref: '#/components/schemas/Parameter'
           type: 'array'
+        supports_image_reference:
+          default: false
+          description: 'Whether this TTS endpoint accepts an `image_url` reference describing the desired voice. Requests carrying an image reference are only routed to endpoints where this is true.'
+          type: 'boolean'
         supports_implicit_caching:
           type: 'boolean'
+        supports_multiple_audio_references:
+          default: false
+          description: 'Whether this TTS endpoint accepts more than one `input_audio` reference clip per request. Requests carrying several clips are only routed to endpoints where this is true.'
+          type: 'boolean'
         supports_tool_choice:
           $ref: '#/components/schemas/ToolChoiceSupport'
         supports_voice_cloning:
@@ -25743,17 +25753,19 @@ components:
         - $ref: '#/components/schemas/ContainerAutoEnvironment'
         - $ref: '#/components/schemas/ContainerReferenceEnvironment'
     SpeechInputReference:
-      description: 'Reference content part for stateless voice cloning'
+      description: 'Reference content part for stateless voice cloning or voice design'
       discriminator:
         mapping:
+          image_url: '#/components/schemas/SpeechInputReferenceImage'
           input_audio: '#/components/schemas/SpeechInputReferenceAudio'
           text: '#/components/schemas/SpeechInputReferenceText'
         propertyName: 'type'
       oneOf:
         - $ref: '#/components/schemas/SpeechInputReferenceAudio'
         - $ref: '#/components/schemas/SpeechInputReferenceText'
+        - $ref: '#/components/schemas/SpeechInputReferenceImage'
     SpeechInputReferenceAudio:
-      description: 'Reference audio input for stateless voice cloning'
+      description: 'Reference audio input for stateless voice cloning. Up to three parts per request; the Nth audio part is addressable from `input` as `@AudioN` on providers that support multiple references.'
       example:
         input_audio:
           data: 'data:audio/wav;base64,UklGRuQXDABXQVZF...'
@@ -25773,7 +25785,7 @@ components:
       description: 'Reference audio input object'
       properties:
         data:
-          description: 'Base64-encoded reference audio (optionally a data URI). Supported audio formats are provider-specific. Limited to 20 MiB of base64 (15 MiB of decoded audio).'
+          description: 'Base64-encoded reference audio (optionally a data URI). Supported audio formats are provider-specific. Limited to 20 MiB of base64 (15 MiB of decoded audio). Exactly one of `data` or `url` is required.'
           example: 'data:audio/wav;base64,UklGRuQXDABXQVZF...'
           maxLength: 20971520
           minLength: 1
@@ -25782,17 +25794,50 @@ components:
           description: 'Audio format of the reference audio (e.g., wav, mp3). Optional; most providers detect the format from the audio bytes.'
           example: 'wav'
           type: 'string'
+        url:
+          description: 'Public http(s) URL of the reference audio. OpenRouter downloads it (15 MiB max) and forwards the bytes, never the URL. Exactly one of `data` or `url` is required.'
+          example: 'https://example.com/reference.wav'
+          format: 'uri'
+          maxLength: 2048
+          type: 'string'
+      type: 'object'
+    SpeechInputReferenceImage:
+      description: 'Reference image describing the desired voice. Cannot be combined with `input_audio` parts. Only routed to endpoints that support image references.'
+      example:
+        image_url:
+          url: 'data:image/png;base64,iVBORw0KGgo...'
+        type: 'image_url'
+      properties:
+        image_url:
+          $ref: '#/components/schemas/SpeechInputReferenceImageInput'
+        type:
+          enum:
+            - 'image_url'
+          type: 'string'
       required:
-        - 'data'
+        - 'type'
+        - 'image_url'
+      type: 'object'
+    SpeechInputReferenceImageInput:
+      description: 'Reference image input object'
+      properties:
+        url:
+          description: 'JPEG, PNG, or WebP reference image as a base64 data URI or a public http(s) URL. Remote images are downloaded (15 MiB max) and forwarded as bytes, never as the URL.'
+          example: 'data:image/png;base64,iVBORw0KGgo...'
+          maxLength: 20971520
+          minLength: 1
+          type: 'string'
+      required:
+        - 'url'
       type: 'object'
     SpeechInputReferenceText:
-      description: 'Transcript of the accompanying reference audio'
+      description: 'Transcript of an `input_audio` part'
       example:
         text: 'I used to rule the world.'
         type: 'text'
       properties:
         text:
-          description: 'Transcript of the accompanying reference audio.'
+          description: 'Transcript of an `input_audio` part. With a single clip it may appear before or after the clip; with multiple clips it must immediately follow the clip it transcribes.'
           example: 'I used to rule the world.'
           maxLength: 10000
           type: 'string'
@@ -25818,7 +25863,7 @@ components:
           example: 'Hello world'
           type: 'string'
         input_references:
-          description: 'Reference content for stateless voice cloning: one `input_audio` part carrying the voice sample, optionally accompanied by one `text` part with its transcript. Only routed to endpoints that support voice cloning.'
+          description: 'Reference content for stateless voice cloning or voice design. Audio mode: one to three `input_audio` parts, each optionally paired with a `text` part carrying its transcript (a single clip accepts its transcript before or after it; with multiple clips each transcript immediately follows its clip); only routed to endpoints that support voice cloning (and multiple references when more than one part is sent). Image mode: exactly one `image_url` part; only routed to endpoints that support image references. The two modes cannot be mixed. An empty array is treated as no reference.'
           example:
             - input_audio:
                 data: 'data:audio/wav;base64,UklGRuQXDABXQVZF...'
@@ -33874,7 +33919,9 @@ paths:
                       - 'temperature'
                       - 'top_p'
                       - 'max_tokens'
+                    supports_image_reference: false
                     supports_implicit_caching: true
+                    supports_multiple_audio_references: false
                     supports_voice_cloning: false
                     tag: 'openai'
                     throughput_last_30m:
@@ -33911,7 +33958,9 @@ paths:
                         - 'temperature'
                         - 'top_p'
                         - 'max_tokens'
+                      supports_image_reference: false
                       supports_implicit_caching: true
+                      supports_multiple_audio_references: false
                       supports_voice_cloning: false
                       tag: 'openai'
                       throughput_last_30m: