feat(anthropic): Add Anthropic PDF support (document type) in invoke

examples/src/prompts/pdf_document.ts

@@ -0,0 +1,48 @@
+import { base } from "@faker-js/faker";
+import { ChatAnthropic } from "@langchain/anthropic";
+
+export const run = async () => {
+  const llm = new ChatAnthropic({
+    model: "claude-3-5-sonnet-20240620", // Only claude-3-5-sonnet-20240620 , claude-3-5-sonnet-20241022 as of Jan 2025 support PDF documents as in base64
+  });
+
+  // PDF needs to be in Base64.
+  const getLocalFile = async (path: string) => {
+    const localFile = await fs.readFileSync(path);
+    const base64File = localFile.toString("base64");
+    return base64File;
+  };
+
+  // Or remotely
+  const getRemoteFile = async (url: string) => {
+    const response = await fetch(url);
+    const arrayBuffer = await response.arrayBuffer();
+    const base64File = Buffer.from(arrayBuffer).toString("base64");
+    return base64File;
+  };
+
+  const base64 = await getRemoteFile(
+    "https://www.w3.org/WAI/ER/tests/xhtml/testfiles/resources/pdf/dummy.pdf"
+  );
+
+  const prompt = "Summarise the contents of this PDF";
+
+  const response = await llm.invoke([
+    {
+      role: "user",
+      content: [
+        {
+          type: "text",
+          text: prompt,
+        },
+        {
+          type: "document",
+          source: base64,
+        },
+      ],
+    },
+  ]);
+  console.log(response.content);
+  //console.log(response.content);
+  return response.content;
+};

libs/langchain-anthropic/src/utils/message_inputs.ts

@@ -131,6 +131,20 @@ function _formatContent(content: MessageContent) {
           source,
           ...(cacheControl ? { cache_control: cacheControl } : {}),
         };
+      } else if (contentPart.type === "document") {
+        // PDF
+        return {
+          type: "document",
+          source:
+            typeof contentPart.source === "string"
+              ? {
+                  media_type: "application/pdf",
+                  type: "base64",
+                  data: contentPart.source,
+                }
+              : contentPart.source,
+          ...(cacheControl ? { cache_control: cacheControl } : {}),
+        };
       } else if (
         textTypes.find((t) => t === contentPart.type) &&
         "text" in contentPart