diff --git a/core/llm/streamChat.ts b/core/llm/streamChat.ts
index ec1f046c5ec..4c70ff990d5 100644
--- a/core/llm/streamChat.ts
+++ b/core/llm/streamChat.ts
@@ -94,13 +94,17 @@ export async function* llmStreamChat(
         completionOptions,
         abortController,
       });
+      let accumulatedCompletion = "";
       let next = await gen.next();
       while (!next.done) {
         if (abortController.signal.aborted) {
+          errorPromptLog.completion = accumulatedCompletion;
           next = await gen.return(errorPromptLog);
           break;
         }
         if (next.value) {
+          const content = typeof next.value === "string" ? next.value : "";
+          accumulatedCompletion += content;
           yield {
             role: "assistant",
             content: next.value,
@@ -120,15 +124,30 @@ export async function* llmStreamChat(
         completionOptions,
         messageOptions,
       );
+      let accumulatedCompletion = "";
       let next = await gen.next();
       while (!next.done) {
         if (abortController.signal.aborted) {
+          errorPromptLog.completion = accumulatedCompletion;
           next = await gen.return(errorPromptLog);
           break;
         }
 
         const chunk = next.value;
 
+        // Accumulate the content from chunks
+        if (chunk.content) {
+          const content =
+            typeof chunk.content === "string"
+              ? chunk.content
+              : Array.isArray(chunk.content)
+                ? chunk.content
+                    .map((part) => (part.type === "text" ? part.text : ""))
+                    .join("")
+                : "";
+          accumulatedCompletion += content;
+        }
+
         yield chunk;
         next = await gen.next();
       }
diff --git a/core/llm/streamChat.vitest.ts b/core/llm/streamChat.vitest.ts
new file mode 100644
index 00000000000..63c0e460824
--- /dev/null
+++ b/core/llm/streamChat.vitest.ts
@@ -0,0 +1,97 @@
+import { describe, test, expect } from "vitest";
+
+/**
+ * Tests for the content accumulation logic used in streamChat.ts.
+ *
+ * The core change in streamChat.ts adds an `accumulatedCompletion` variable
+ * that tracks partial output from streaming chunks. This logic must handle
+ * both string content and MessagePart[] content correctly.
+ *
+ * These are unit tests for the extraction/accumulation behavior.
+ * Integration tests for the full llmStreamChat flow are covered by
+ * existing e2e tests.
+ */
+describe("streamChat content accumulation logic", () => {
+  // Mirror the extraction logic from streamChat.ts lines 140-147
+  function extractContent(content: unknown): string {
+    if (typeof content === "string") {
+      return content;
+    }
+    if (Array.isArray(content)) {
+      return content
+        .map((part: any) => (part.type === "text" ? part.text : ""))
+        .join("");
+    }
+    return "";
+  }
+
+  test("should extract string content from chunks", () => {
+    expect(extractContent("Hello world")).toBe("Hello world");
+  });
+
+  test("should extract text from MessagePart[] content", () => {
+    const parts = [
+      { type: "text", text: "Part 1 " },
+      { type: "text", text: "Part 2" },
+    ];
+    expect(extractContent(parts)).toBe("Part 1 Part 2");
+  });
+
+  test("should skip non-text MessageParts (e.g. imageUrl)", () => {
+    const parts = [
+      { type: "text", text: "Hello " },
+      { type: "imageUrl", imageUrl: { url: "http://example.com/img.png" } },
+      { type: "text", text: "world" },
+    ];
+    expect(extractContent(parts)).toBe("Hello world");
+  });
+
+  test("should return empty string for undefined/null content", () => {
+    expect(extractContent(undefined)).toBe("");
+    expect(extractContent(null)).toBe("");
+  });
+
+  test("should accumulate content across multiple streaming chunks", () => {
+    const chunks = [
+      { content: "Hello " },
+      { content: "world" },
+      { content: "!" },
+    ];
+    let accumulated = "";
+    for (const chunk of chunks) {
+      accumulated += extractContent(chunk.content);
+    }
+    expect(accumulated).toBe("Hello world!");
+  });
+
+  test("should accumulate mixed string and MessagePart[] chunks", () => {
+    const chunks = [
+      { content: "Start " },
+      {
+        content: [
+          { type: "text", text: "middle " },
+          { type: "text", text: "part" },
+        ],
+      },
+      { content: " end" },
+    ];
+    let accumulated = "";
+    for (const chunk of chunks) {
+      accumulated += extractContent(chunk.content);
+    }
+    expect(accumulated).toBe("Start middle part end");
+  });
+
+  test("should handle empty chunks without error", () => {
+    const chunks = [
+      { content: "Hello" },
+      { content: "" },
+      { content: " world" },
+    ];
+    let accumulated = "";
+    for (const chunk of chunks) {
+      accumulated += extractContent(chunk.content);
+    }
+    expect(accumulated).toBe("Hello world");
+  });
+});