feat: enhance audio, image, and video asset schemas with new properti…

schemas/audioasset.yaml

@@ -53,10 +53,32 @@
           default: false
         model:
           description: >-
-            The generation model to use when `prompt` is set (e.g. `polly-neural`).
+            The generation model to use when `prompt` is set (e.g. `polly-neural`,
+            `fal/elevenlabs-tts`, `fal/minimax-speech`, `fal/elevenlabs-music`).
             Defaults to the platform's preferred generator if omitted.
           type: string
           example: polly-neural
+        musicLengthMs:
+          description: >-
+            Target duration in milliseconds for generated music (3,000–600,000).
+            Only meaningful when `prompt` is set and the model is a music generator
+            (e.g. `fal/elevenlabs-music`).
+          type: number
+          minimum: 3000
+          maximum: 600000
+          example: 30000
+        forceInstrumental:
+          description: >-
+            Set to `true` to force instrumental music (no vocals). Only meaningful
+            for music generation models.
+          type: boolean
+          default: false
+        compositionPlan:
+          description: >-
+            An optional structured composition plan for music generation. Only
+            meaningful for music generation models (e.g. `fal/elevenlabs-music`).
+          type: string
+          maxLength: 4000
         trim:
           description: >-
             The start trim point of the audio clip, in seconds (defaults to 0). Audio

schemas/imageasset.yaml

@@ -32,10 +32,35 @@
           example: A serene landscape with a crystal-clear mountain lake at sunrise.
         model:
           description: >-
-            The generation model to use when `prompt` is set (e.g. `flux-schnell`).
-            Defaults to the platform's preferred generator if omitted.
+            The generation model to use when `prompt` is set (e.g. `flux-schnell`,
+            `fal/flux-schnell`, `fal/nano-banana-2`). Defaults to the platform's
+            preferred generator if omitted.
           type: string
           example: flux-schnell
+        resolution:
+          description: >-
+            Output resolution tier for supported image generation models.
+            `1K` (default), `2K`, or `4K`. Only meaningful when `prompt` is set.
+          enum:
+            - '1K'
+            - '2K'
+            - '4K'
+          type: string
+          example: 1K
+        aspectRatio:
+          description: >-
+            Aspect ratio for the generated image. Only meaningful when `prompt`
+            is set and the model supports it.
+          enum:
+            - '1:1'
+            - '16:9'
+            - '9:16'
+            - '4:3'
+            - '3:4'
+            - '21:9'
+            - '9:21'
+          type: string
+          example: '1:1'
         crop:
           $ref: "./crop.yaml#/Crop"
       additionalProperties: false

schemas/videoasset.yaml

@@ -43,10 +43,63 @@
         model:
           description: >-
             The generation model to use when `prompt` is set (e.g. `luma-ray-3`,
-            `runpod-itv-mini`). Defaults to the platform's preferred generator if
-            omitted.
+            `runpod-itv-mini`, `fal/seedance-2.0`). Defaults to the platform's
+            preferred generator if omitted.
           type: string
           example: luma-ray-3
+        resolution:
+          description: >-
+            Output resolution for video generation. Only meaningful when `prompt`
+            is set and the model supports it (e.g. `fal/seedance-2.0`).
+          enum:
+            - '480p'
+            - '720p'
+            - '1080p'
+          type: string
+          example: '720p'
+        duration:
+          description: >-
+            Target video duration in seconds for generation models that accept
+            it. `"auto"` lets the model decide. Only meaningful when `prompt`
+            is set.
+          enum:
+            - 'auto'
+            - '4'
+            - '5'
+            - '6'
+            - '7'
+            - '8'
+            - '9'
+            - '10'
+            - '11'
+            - '12'
+            - '13'
+            - '14'
+            - '15'
+          type: string
+          default: auto
+          example: '5'
+        aspectRatio:
+          description: >-
+            Aspect ratio for the generated video. Only meaningful when `prompt`
+            is set and the model supports it.
+          enum:
+            - '1:1'
+            - '16:9'
+            - '9:16'
+            - '4:3'
+            - '3:4'
+            - '21:9'
+            - '9:21'
+          type: string
+          example: '16:9'
+        generateAudio:
+          description: >-
+            Set to `true` to request audio generation alongside the video. Only
+            meaningful for video generation models that support it (e.g.
+            `fal/seedance-2.0`).
+          type: boolean
+          default: false
         transcode:
           type: boolean
           description: >-