diff --git a/aborist/qa/query.py b/aborist/qa/query.py
index ad5e7d1..308d0f8 100644
--- a/aborist/qa/query.py
+++ b/aborist/qa/query.py
@@ -309,16 +309,25 @@ def query(
 ) -> dict:
     """Answer `question` using the corpus. Cache to qa_db. Returns a result dict."""
     policy = policy or DEFAULT_QUERY_POLICY
+    t_start = time.monotonic()
 
     # 1. Search.
+    t_phase = time.monotonic()
     hits = _search_corpus(shards_dir, single_db, question, over_fetch)
     if not hits:
-        return {"status": "no_sources", "msg": "FTS5 search returned no hits"}
-    # 1b. Rerank — title-boost gives topically-named docs priority over
-    # body-only matches like Tell_(poker) for "tell me about ...".
+        return {
+            "status": "no_sources",
+            "msg": "FTS5 search returned no hits",
+            "timings": {
+                "search_ms": _ms_since(t_phase),
+                "total_ms": _ms_since(t_start),
+            },
+        }
     hits = _rerank(hits, question)
+    search_ms = _ms_since(t_phase)
 
     # 2. Pull doc texts within budget.
+    t_phase = time.monotonic()
     chosen: list[_Hit] = []
     context_parts: list[str] = []
     char_budget = max_context_chars
@@ -337,9 +346,17 @@ def query(
             break
 
     if not chosen:
-        return {"status": "no_sources", "msg": "top-k hits had cold or empty content"}
-
+        return {
+            "status": "no_sources",
+            "msg": "top-k hits had cold or empty content",
+            "timings": {
+                "search_ms": search_ms,
+                "context_ms": _ms_since(t_phase),
+                "total_ms": _ms_since(t_start),
+            },
+        }
     context = "\n\n".join(context_parts)
+    context_ms = _ms_since(t_phase)
 
     # 3. Build messages + hashes.
     messages = [
@@ -368,11 +385,13 @@ def query(
     qa_conn = connect(qa_db)
     try:
         # 4. Cache lookup.
+        t_phase = time.monotonic()
         cached = qa_conn.execute(
             "SELECT * FROM providence_cache "
             "WHERE cache_key = ? AND falsification_state = 'live'",
             (ckey,),
         ).fetchone()
+        cache_lookup_ms = _ms_since(t_phase)
         if cached is not None:
             now = int(time.time())
             with transaction(qa_conn):
@@ -389,9 +408,18 @@ def query(
                 "context_root": context_root,
                 "answer_text": cached["answer_text"],
                 "sources": json.loads(cached["merkle_proof"])["sources"],
+                "timings": {
+                    "search_ms": search_ms,
+                    "context_ms": context_ms,
+                    "cache_lookup_ms": cache_lookup_ms,
+                    "llm_ms": None,
+                    "persist_ms": None,
+                    "total_ms": _ms_since(t_start),
+                },
             }
 
         # 5. Cache miss — call LLM.
+        t_phase = time.monotonic()
         answer_text = chat_client.chat_completion(
             messages,
             model=model_id,
@@ -399,8 +427,10 @@ def query(
             max_tokens=policy["max_tokens"],
             top_p=policy.get("top_p", 1.0),
         )
+        llm_ms = _ms_since(t_phase)
 
         # 6. Persist record + audit event.
+        t_phase = time.monotonic()
         proof_obj = {
             "context_root": context_root,
             "sources": [
@@ -463,6 +493,7 @@ def query(
     finally:
         qa_conn.close()
 
+    persist_ms = _ms_since(t_phase)
     return {
         "status": "cache_miss_then_written",
         "audit_mode": "STRICT",
@@ -470,4 +501,17 @@ def query(
         "context_root": context_root,
         "answer_text": answer_text,
         "sources": proof_obj["sources"],
+        "timings": {
+            "search_ms": search_ms,
+            "context_ms": context_ms,
+            "cache_lookup_ms": cache_lookup_ms,
+            "llm_ms": llm_ms,
+            "persist_ms": persist_ms,
+            "total_ms": _ms_since(t_start),
+        },
     }
+
+
+def _ms_since(t: float) -> float:
+    """Wall-time elapsed in milliseconds, rounded to 1 decimal."""
+    return round((time.monotonic() - t) * 1000, 1)
diff --git a/aborist/qa/runner.py b/aborist/qa/runner.py
index 49a4ef6..e0a1877 100644
--- a/aborist/qa/runner.py
+++ b/aborist/qa/runner.py
@@ -42,6 +42,10 @@ DEFAULT_POLICY = {
 }
 
 
+def _ms_since(t: float) -> float:
+    return round((time.monotonic() - t) * 1000, 1)
+
+
 def ask(
     conn: sqlite3.Connection,
     *,
@@ -56,6 +60,7 @@ def ask(
 ) -> dict:
     """Look up cached answer or run inference. Returns a result dict."""
     policy = policy or DEFAULT_POLICY
+    t_start = time.monotonic()
 
     doc = conn.execute(
         "SELECT document_uri, chunking_version FROM documents "
@@ -103,11 +108,13 @@ def ask(
         doc["chunking_version"],
     )
 
+    t_lookup = time.monotonic()
     cached = conn.execute(
         "SELECT * FROM providence_cache "
         "WHERE cache_key = ? AND falsification_state = 'live'",
         (ckey,),
     ).fetchone()
+    cache_lookup_ms = _ms_since(t_lookup)
     if cached is not None:
         with transaction(conn):
             now = int(time.time())
@@ -124,8 +131,14 @@ def ask(
             "source_root": document_root,
             "answer_text": cached["answer_text"],
             "merkle_proof": json.loads(cached["merkle_proof"]),
+            "timings": {
+                "cache_lookup_ms": cache_lookup_ms,
+                "llm_ms": None,
+                "total_ms": _ms_since(t_start),
+            },
         }
 
+    t_llm = time.monotonic()
     answer_text = client.chat_completion(
         messages,
         model=model_id,
@@ -133,6 +146,7 @@ def ask(
         max_tokens=policy["max_tokens"],
         top_p=policy.get("top_p", 1.0),
     )
+    llm_ms = _ms_since(t_llm)
 
     leaves = [bytes.fromhex(r["leaf_hash"]) for r in chunk_rows]
     tree = MerkleTree.build(leaves)
@@ -193,4 +207,9 @@ def ask(
         "source_root": document_root,
         "answer_text": answer_text,
         "merkle_proof": proof_obj,
+        "timings": {
+            "cache_lookup_ms": cache_lookup_ms,
+            "llm_ms": llm_ms,
+            "total_ms": _ms_since(t_start),
+        },
     }