diff --git a/packages/ai/src/api/pi-messages.ts b/packages/ai/src/api/pi-messages.ts
index 605194937..83ddc06c7 100644
--- a/packages/ai/src/api/pi-messages.ts
+++ b/packages/ai/src/api/pi-messages.ts
@@ -74,6 +74,7 @@ export type PiMessagesEvent =
 			reason: Extract<PiMessagesStopReason, "stop" | "length" | "toolUse">;
 			usage: PiMessagesUsage;
 			responseId?: string;
+			responseModel?: string;
 			providerThinkingLevel?: string;
 			rewrite?: PiMessagesRewriteImpact;
 	  }
@@ -83,6 +84,7 @@ export type PiMessagesEvent =
 			usage: PiMessagesUsage;
 			errorMessage?: string;
 			responseId?: string;
+			responseModel?: string;
 			providerThinkingLevel?: string;
 			rewrite?: PiMessagesRewriteImpact;
 	  };
@@ -199,6 +201,9 @@ function createEventConverter(model: Model<"pi-messages">) {
 					usage: event.usage,
 					responseId: event.responseId,
 				});
+				if (event.responseModel !== undefined) {
+					partial.responseModel = event.responseModel;
+				}
 				if (event.providerThinkingLevel !== undefined) {
 					partial.providerThinkingLevel = event.providerThinkingLevel;
 				}
@@ -211,6 +216,9 @@ function createEventConverter(model: Model<"pi-messages">) {
 					errorMessage: event.errorMessage,
 					responseId: event.responseId,
 				});
+				if (event.responseModel !== undefined) {
+					partial.responseModel = event.responseModel;
+				}
 				if (event.providerThinkingLevel !== undefined) {
 					partial.providerThinkingLevel = event.providerThinkingLevel;
 				}
diff --git a/packages/ai/src/types.ts b/packages/ai/src/types.ts
index d1fcee314..32b5a5f31 100644
--- a/packages/ai/src/types.ts
+++ b/packages/ai/src/types.ts
@@ -588,7 +588,7 @@ export interface AssistantMessage {
 	api: Api;
 	provider: ProviderId;
 	model: string;
-	responseModel?: string; // Concrete model reported by the provider when different from the requested `model`
+	responseModel?: string; // Concrete model reported by the provider, when reported. May equal the requested `model`
 	responseId?: string; // Provider-specific response/message identifier when the upstream API exposes one
 	/** Exact provider-native effort level used for this response. Absent for legacy or unmanaged responses. */
 	providerThinkingLevel?: string;
diff --git a/packages/ai/test/pi-messages.test.ts b/packages/ai/test/pi-messages.test.ts
index 211c24cb0..54b3d4d9e 100644
--- a/packages/ai/test/pi-messages.test.ts
+++ b/packages/ai/test/pi-messages.test.ts
@@ -116,6 +116,7 @@ describe("pi-messages", () => {
 					reason: "toolUse",
 					usage,
 					responseId: "resp_1",
+					responseModel: "claude-opus-5-5",
 					providerThinkingLevel: "high",
 				},
 			],
@@ -145,6 +146,7 @@ describe("pi-messages", () => {
 		expect(message.responseId).toBe("resp_1");
 		expect(message.providerThinkingLevel).toBe("high");
 		expect(message.model).toBe("auto");
+		expect(message.responseModel).toBe("claude-opus-5-5");
 		expect(message.provider).toBe("radius");
 		expect(message.content).toEqual([
 			{ type: "text", text: "Hello", textSignature: undefined },
@@ -190,6 +192,7 @@ describe("pi-messages", () => {
 		expect(eventModels).toEqual(wireEvents.map(() => model));
 		expect(message.stopReason).toBe("stop");
 		expect(message.responseId).toBe("resp_1");
+		expect(message).not.toHaveProperty("responseModel");
 		expect(message.content).toEqual([{ type: "text", text: "Hello", textSignature: undefined }]);
 	});
 
@@ -234,7 +237,16 @@ describe("pi-messages", () => {
 
 	it("propagates server-sent error events", async () => {
 		const { baseUrl } = await startServer({
-			events: [{ type: "start" }, { type: "error", reason: "error", usage, errorMessage: "Upstream failed" }],
+			events: [
+				{ type: "start" },
+				{
+					type: "error",
+					reason: "error",
+					usage,
+					errorMessage: "Upstream failed",
+					responseModel: "claude-opus-5-5",
+				},
+			],
 		});
 		const model = createModel(baseUrl);
 
@@ -243,6 +255,7 @@ describe("pi-messages", () => {
 		expect(message.stopReason).toBe("error");
 		expect(message.errorMessage).toBe("Upstream failed");
 		expect(message.usage).toEqual(usage);
+		expect(message.responseModel).toBe("claude-opus-5-5");
 	});
 
 	it("errors when no API key is provided", async () => {
diff --git a/packages/coding-agent/docs/message-types.md b/packages/coding-agent/docs/message-types.md
index dfdafd505..6c5d95c63 100644
--- a/packages/coding-agent/docs/message-types.md
+++ b/packages/coding-agent/docs/message-types.md
@@ -140,7 +140,7 @@ interface AssistantMessage {
 }
 ```
 
-`responseModel` records a concrete provider response model when it differs from the requested model. `responseId`, `providerThinkingLevel`, `thinkingLevel`, `diagnostics`, and `rawStopReason` preserve provider or runtime details.
+`responseModel` records the concrete model reported by the provider, when the provider reports one. It may equal the requested model. `responseId`, `providerThinkingLevel`, `thinkingLevel`, `diagnostics`, and `rawStopReason` preserve provider or runtime details.
 
 `"pending"` is used for a partial assistant message while it streams. The completed message in `message_end` has a terminal stop reason, and Pi does not persist `"pending"` assistant messages in session JSONL.