docs: clarify deepseek token limit source

frontend/src/lib/components/copilot/lib.ts

@@ -315,7 +315,8 @@ export function getModelMaxTokens(provider: AIProvider, model: string) {
 		provider === 'deepseek' &&
 		/^(deepseek-(flash|chat|reasoner)|deepseek-v4-(flash|pro))$/.test(parseModelId(model).base)
 	) {
-		// DeepSeek counts reasoning against the output budget as well as the answer.
+		// DeepSeek's documented 384K output ceiling includes reasoning as well as the answer.
+		// https://api-docs.deepseek.com/api/create-chat-completion/
 		return 393216
 	} else if (model.includes('gpt-5')) {
 		return 128000

frontend/src/lib/components/copilot/modelConfig.ts

@@ -96,7 +96,7 @@ const MODEL_CONTEXT_WINDOWS: [name: string, contextWindow: number][] = [
 	['gemini-3.1', 1_000_000],
 	['gemini-3', 1_000_000],
 	['gemini-2.5', 1_000_000],
-	// DeepSeek — the V4 family (pro / flash) is 1M. The deepseek-chat /
+	// DeepSeek — Flash and the V4 family (pro / flash) are 1M. The deepseek-chat /
 	// deepseek-reasoner aliases were retired 2026-07-24 but can still sit in a
 	// saved selection, so they keep resolving to the window they had.
 	['deepseek-v4', 1_000_000],