@shop-os/modelkit 0.2.15 → 0.2.17

package/CHANGELOG.md

@@ -9,6 +9,19 @@ header in the release PR.
 
 ## [Unreleased]
 
+## [0.2.17] - Unreleased
+
+### Fixed
+- Gemini Omni extension: an extension started from an uploaded source no longer deletes the upload once the interaction is created. The stored interaction keeps referencing that file, so a follow-up call chaining on it by `previousInteractionId` failed with `403 You do not have permission to access the content` (only the first of a multi-segment extension ever ran). Uploads for extend-from-source calls are left to the Files API's 48 h expiry; every other call still cleans up immediately.
+
+## [0.2.16] - 2026-09-24
+
+### Added
+- Gemini Omni 1.1 Flash video extension. The row declares `supports: ["videoExtension"]` and `constraints.extension: { segmentSeconds: 10, maxTotalSeconds: 40 }`. A `taskType: "extend"` request continues the clip by `durationSeconds` and the result is the **full cumulative clip** (probe 2026-09-24). The first extension may upload a source video (≤ 10 s, `sourceVideo`); follow-ups chain with `VideoRequest.previousInteractionId`, taken from the previous result's new `VideoResult.providerRef.interactionId`. The Interactions API rejects `previous_interaction_id` together with a video task, so chained calls carry only the prompt (no `generation_config`) and no aspect ratio.
+- `SupportedOption` / catalog `supports` gain `"videoExtension"`; `ModelConstraints.extension` is typed and validated.
+- A chained extension (`previousInteractionId` set) sends the prompt alone: `referenceImages`, `referenceVideos` and `firstFrame` are dropped with an `option_ignored` warning each, since the API rejects the id next to a video task.
+- `resolveVideoOptions` keeps `taskType: "extend"` on a `videoExtension` row (it is otherwise dropped as unsupported) and drops `previousInteractionId` with an `option_ignored` warning on rows without the capability.
+
 ## [0.2.15] - Unreleased
 
 ### Added

package/dist/catalog.cjs

@@ -1446,7 +1446,14 @@ var VIDEO_MODELS = [
       durations: [3, 4, 5, 6, 7, 8, 9, 10],
       references: { images: 6, videos: 1, videoSeconds: { max: 10 } },
       sourceVideoSeconds: { max: 10 },
-      supports: []
+      // Extension (probe 2026-09-24, 360p): each `taskType: "extend"` call
+      // renders `durationSeconds` more and returns the FULL cumulative clip
+      // (3 s -> 6 s -> 9 s); the provider caps the clip at 40 s. A first
+      // extension can upload a <= 10 s source with task "extend"; later ones
+      // chain by `previousInteractionId` (the API rejects the id together
+      // with a video task, so those calls carry the prompt alone).
+      supports: ["videoExtension"],
+      extension: { segmentSeconds: 10, maxTotalSeconds: 40 }
     },
     defaults: {},
     usage: { ...VIDEO_USAGE },
@@ -1454,7 +1461,7 @@ var VIDEO_MODELS = [
     display: {
       provider: "google",
       logo: GOOGLE_LOGO2,
-      description: "Google Gemini Omni Flash 1.1 conversational video generation and editing; text, image, and up to 10-second video input, 3-10 second output at up to 4K and 24 FPS."
+      description: "Google Gemini Omni Flash 1.1 conversational video generation, editing and extension; text, image, and up to 10-second video input, 3-10 second output at up to 4K and 24 FPS, extendable to 40 seconds."
     },
     capabilities: { acceptsSourceVideo: true }
   },
@@ -2185,6 +2192,7 @@ var SupportedOptionSchema = import_zod2.z.enum([
   "watermark",
   "safetyChecker",
   "taskType",
+  "videoExtension",
   "instrumental"
 ]);
 var UsageUnitSchema = import_zod2.z.enum([
@@ -2232,7 +2240,8 @@ var ModelConstraintsSchema = import_zod2.z.looseObject({
   requires: import_zod2.z.array(RequiredInputSchema).optional(),
   supports: import_zod2.z.array(SupportedOptionSchema).optional(),
   ratioIgnoredWith: import_zod2.z.array(import_zod2.z.enum(["firstFrame", "sourceVideo"])).optional(),
-  sourceVideoSeconds: VideoSecondsLimitsSchema.omit({ total: true }).extend({ inReferenceTotal: import_zod2.z.boolean().optional() }).optional()
+  sourceVideoSeconds: VideoSecondsLimitsSchema.omit({ total: true }).extend({ inReferenceTotal: import_zod2.z.boolean().optional() }).optional(),
+  extension: import_zod2.z.looseObject({ segmentSeconds: import_zod2.z.number(), maxTotalSeconds: import_zod2.z.number() }).optional()
 });
 var MediaRefSchema = import_zod2.z.union([
   import_zod2.z.looseObject({
@@ -2979,6 +2988,7 @@ function resolveVideoOptions(model, request, onUnsupported = "error") {
   setOrDelete(merged, "referenceImages", images);
   setOrDelete(merged, "referenceVideos", videos);
   setOrDelete(merged, "referenceAudios", audios);
+  const keepExtend = merged.taskType === "extend" && (model.constraints.supports ?? []).includes("videoExtension");
   dropUnsupported(merged, model, warnings, [
     "seed",
     "negativePrompt",
@@ -2988,8 +2998,17 @@ function resolveVideoOptions(model, request, onUnsupported = "error") {
     "lastFrame",
     "watermark",
     "safetyChecker",
-    "taskType"
+    ...keepExtend ? [] : ["taskType"]
   ]);
+  if (merged.previousInteractionId !== void 0 && !keepExtend) {
+    warnings.push({
+      code: "option_ignored",
+      field: "previousInteractionId",
+      from: merged.previousInteractionId,
+      message: `${model.id} does not support previousInteractionId; ignored`
+    });
+    delete merged.previousInteractionId;
+  }
   return { request: merged, warnings };
 }
 function resolveAudioOptions(model, request, onUnsupported = "error") {