feat(llm): resolve models as config > env > default

Separate hardcoded default from env var in DEFAULT_EMBED_MODEL so the
constructor can resolve: config param > env var > hardcoded default.
Also add env var support for QMD_GENERATE_MODEL and QMD_RERANK_MODEL.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

src/llm.ts

@@ -193,7 +193,7 @@ export type RerankDocument = {
 // HuggingFace model URIs for node-llama-cpp
 // Format: hf:<user>/<repo>/<file>
 // Override via QMD_EMBED_MODEL env var (e.g. hf:Qwen/Qwen3-Embedding-0.6B-GGUF/Qwen3-Embedding-0.6B-Q8_0.gguf)
-const DEFAULT_EMBED_MODEL = process.env.QMD_EMBED_MODEL ?? "hf:ggml-org/embeddinggemma-300M-GGUF/embeddinggemma-300M-Q8_0.gguf";
+const DEFAULT_EMBED_MODEL = "hf:ggml-org/embeddinggemma-300M-GGUF/embeddinggemma-300M-Q8_0.gguf";
 const DEFAULT_RERANK_MODEL = "hf:ggml-org/Qwen3-Reranker-0.6B-Q8_0-GGUF/qwen3-reranker-0.6b-q8_0.gguf";
 // const DEFAULT_GENERATE_MODEL = "hf:ggml-org/Qwen3-0.6B-GGUF/Qwen3-0.6B-Q8_0.gguf";
 const DEFAULT_GENERATE_MODEL = "hf:tobil/qmd-query-expansion-1.7B-gguf/qmd-query-expansion-1.7B-q4_k_m.gguf";
@@ -436,9 +436,9 @@ export class LlamaCpp implements LLM {
 
 
   constructor(config: LlamaCppConfig = {}) {
-    this.embedModelUri = config.embedModel || DEFAULT_EMBED_MODEL;
-    this.generateModelUri = config.generateModel || DEFAULT_GENERATE_MODEL;
-    this.rerankModelUri = config.rerankModel || DEFAULT_RERANK_MODEL;
+    this.embedModelUri = config.embedModel || process.env.QMD_EMBED_MODEL || DEFAULT_EMBED_MODEL;
+    this.generateModelUri = config.generateModel || process.env.QMD_GENERATE_MODEL || DEFAULT_GENERATE_MODEL;
+    this.rerankModelUri = config.rerankModel || process.env.QMD_RERANK_MODEL || DEFAULT_RERANK_MODEL;
     this.modelCacheDir = config.modelCacheDir || MODEL_CACHE_DIR;
     this.expandContextSize = resolveExpandContextSize(config.expandContextSize);
     this.inactivityTimeoutMs = config.inactivityTimeoutMs ?? DEFAULT_INACTIVITY_TIMEOUT_MS;
@@ -1559,8 +1559,7 @@ let defaultLlamaCpp: LlamaCpp | null = null;
  */
 export function getDefaultLlamaCpp(): LlamaCpp {
   if (!defaultLlamaCpp) {
-    const embedModel = process.env.QMD_EMBED_MODEL;
-    defaultLlamaCpp = new LlamaCpp(embedModel ? { embedModel } : {});
+    defaultLlamaCpp = new LlamaCpp();
   }
   return defaultLlamaCpp;
 }

test/llm.test.ts

@@ -117,6 +117,50 @@ describe("LlamaCpp expand context size config", () => {
   });
 });
 
+describe("LlamaCpp model resolution (config > env > default)", () => {
+  const HARDCODED_EMBED = "hf:ggml-org/embeddinggemma-300M-GGUF/embeddinggemma-300M-Q8_0.gguf";
+  const HARDCODED_RERANK = "hf:ggml-org/Qwen3-Reranker-0.6B-Q8_0-GGUF/qwen3-reranker-0.6b-q8_0.gguf";
+  const HARDCODED_GENERATE = "hf:tobil/qmd-query-expansion-1.7B-gguf/qmd-query-expansion-1.7B-q4_k_m.gguf";
+
+  test("uses hardcoded default when no config or env is set", () => {
+    const prev = process.env.QMD_EMBED_MODEL;
+    delete process.env.QMD_EMBED_MODEL;
+    try {
+      const llm = new LlamaCpp({}) as any;
+      expect(llm.embedModelUri).toBe(HARDCODED_EMBED);
+      expect(llm.rerankModelUri).toBe(HARDCODED_RERANK);
+      expect(llm.generateModelUri).toBe(HARDCODED_GENERATE);
+    } finally {
+      if (prev === undefined) delete process.env.QMD_EMBED_MODEL;
+      else process.env.QMD_EMBED_MODEL = prev;
+    }
+  });
+
+  test("env var overrides hardcoded default", () => {
+    const prev = process.env.QMD_EMBED_MODEL;
+    process.env.QMD_EMBED_MODEL = "hf:custom/embed-model.gguf";
+    try {
+      const llm = new LlamaCpp({}) as any;
+      expect(llm.embedModelUri).toBe("hf:custom/embed-model.gguf");
+    } finally {
+      if (prev === undefined) delete process.env.QMD_EMBED_MODEL;
+      else process.env.QMD_EMBED_MODEL = prev;
+    }
+  });
+
+  test("config overrides env var", () => {
+    const prev = process.env.QMD_EMBED_MODEL;
+    process.env.QMD_EMBED_MODEL = "hf:env/model.gguf";
+    try {
+      const llm = new LlamaCpp({ embedModel: "hf:config/model.gguf" }) as any;
+      expect(llm.embedModelUri).toBe("hf:config/model.gguf");
+    } finally {
+      if (prev === undefined) delete process.env.QMD_EMBED_MODEL;
+      else process.env.QMD_EMBED_MODEL = prev;
+    }
+  });
+});
+
 describe("LlamaCpp rerank deduping", () => {
   test("deduplicates identical document texts before scoring", async () => {
     const llm = new LlamaCpp({}) as any;