Add local/lorem-ipsum model to avoid OpenRouter spend

pointer/config/models.yaml

@@ -1,11 +1,14 @@
 # OpenRouter model slugs (https://openrouter.ai/api/v1/models). `*` grants any
 # slug; a `:*` suffix grants every variant of one model, e.g. `:batch`.
 basic:
+  - local/lorem-ipsum
   - poolside/laguna-s-2.1:free
 advanced:
+  - local/lorem-ipsum
   - poolside/laguna-s-2.1:free
   - nvidia/nemotron-3.5-lightning:free
 premium:
+  - local/lorem-ipsum
   - poolside/laguna-s-2.1:free
   - nvidia/nemotron-3.5-lightning:free
   - google/gemma-4-31b-it:free

pointer/lib/generate/index.ts

@@ -1,6 +1,7 @@
 import { streamText } from "ai";
 
 import { chatModel } from "./openrouter";
+import { LOREM_MODEL, streamLoremIpsum } from "./lorem";
 
 const MAX_PROMPT_CHARS = 8_000;
 const MAX_OUTPUT_TOKENS = 4_096;
@@ -98,6 +99,10 @@ export function streamGeneration(
 	input: GenerateInput,
 	signal: AbortSignal,
 ): GenerationStream {
+	if (input.model === LOREM_MODEL) {
+		return streamLoremIpsum(input, signal);
+	}
+
 	let failure: unknown;
 
 	const result = streamText({

pointer/lib/generate/lorem.ts

@@ -0,0 +1,51 @@
+import type { GenerateInput, Generation, GenerationStream } from "./index";
+
+// Must match the slug added to config/models.yaml.
+export const LOREM_MODEL = "local/lorem-ipsum";
+
+const LOREM_TEXT =
+	"Lorem ipsum dolor sit amet, consectetur adipiscing elit. " +
+	"Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. " +
+	"Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris " +
+	"nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in " +
+	"reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla " +
+	"pariatur. Excepteur sint occaecat cupidatat non proident, sunt in " +
+	"culpa qui officia deserunt mollit anim id est laborum.";
+
+// Simulates realistic streaming cadence without hitting the network.
+const CHUNK_DELAY_MS = 40;
+
+// Same ratio as estimateTokens() in index.ts — kept local to avoid a circular import.
+const estimate = (text: string) => Math.max(1, Math.ceil(text.length / 4));
+
+/**
+ * Streams lorem ipsum word-by-word without calling any external API.
+ * Drop-in replacement for streamGeneration() when model === LOREM_MODEL.
+ */
+export function streamLoremIpsum(
+	input: GenerateInput,
+	signal: AbortSignal,
+): GenerationStream {
+	const words = LOREM_TEXT.split(" ");
+	let streamed = "";
+
+	async function* read(): AsyncGenerator<string> {
+		for (const word of words) {
+			if (signal.aborted) { break; }
+			const chunk = streamed.length === 0 ? word : ` ${word}`;
+			streamed += chunk;
+			yield chunk;
+			await new Promise<void>((resolve) => setTimeout(resolve, CHUNK_DELAY_MS));
+		}
+	}
+
+	return {
+		deltas: read(),
+		streamedTokens: () => estimate(streamed),
+		settle: async (): Promise<Generation> => ({
+			output: streamed,
+			inputTokens: estimate(input.prompt),
+			outputTokens: estimate(streamed),
+		}),
+	};
+}