feat (provider/google): add cachedContent optional setting (#2261)

vercel · Jul 12, 2024 · 2e59d26 · 2e59d26
1 parent 4dfe0b0
commit 2e59d26
Show file tree

Hide file tree

Showing 4 changed files with 20 additions and 0 deletions.
diff --git a/.changeset/cool-donkeys-build.md b/.changeset/cool-donkeys-build.md
@@ -0,0 +1,5 @@
+---
+'@ai-sdk/google': patch
+---
+
+feat (provider/google): add cachedContent optional setting
diff --git a/content/providers/01-ai-sdk-providers/10-google-generative-ai.mdx b/content/providers/01-ai-sdk-providers/10-google-generative-ai.mdx
@@ -93,6 +93,11 @@ The following optional settings are available for Google Generative AI models:
   Top-k sampling considers the set of topK most probable tokens.
   Models running with nucleus sampling don't allow topK setting.
 
+- **cachedContent** _string_
+
+  Optional. The name of the cached content used as context to serve the prediction.
+  Format: cachedContents/{cachedContent}
+
 - **safetySettings** _Array\<\{ category: string; threshold: string \}\>_
 
   Optional. Safety settings for the model.

diff --git a/packages/google/src/google-generative-ai-language-model.ts b/packages/google/src/google-generative-ai-language-model.ts
@@ -110,6 +110,7 @@ export class GoogleGenerativeAILanguageModel implements LanguageModelV1 {
             systemInstruction,
             safetySettings: this.settings.safetySettings,
             ...prepareToolsAndToolConfig(mode),
+            cachedContent: this.settings.cachedContent,
           },
           warnings,
         };
@@ -125,6 +126,7 @@ export class GoogleGenerativeAILanguageModel implements LanguageModelV1 {
             contents,
             systemInstruction,
             safetySettings: this.settings.safetySettings,
+            cachedContent: this.settings.cachedContent,
           },
           warnings,
         };
@@ -146,6 +148,7 @@ export class GoogleGenerativeAILanguageModel implements LanguageModelV1 {
             },
             toolConfig: { functionCallingConfig: { mode: 'ANY' } },
             safetySettings: this.settings.safetySettings,
+            cachedContent: this.settings.cachedContent,
           },
           warnings,
         };

diff --git a/packages/google/src/google-generative-ai-settings.ts b/packages/google/src/google-generative-ai-settings.ts
@@ -16,6 +16,13 @@ Models running with nucleus sampling don't allow topK setting.
    */
   topK?: number;
 
+  /**
+Optional.
+The name of the cached content used as context to serve the prediction.
+Format: cachedContents/{cachedContent}
+   */
+  cachedContent?: string;
+
   /**
 Optional. A list of unique safety settings for blocking unsafe content.
    */