diff --git a/agent/prompt_builder.py b/agent/prompt_builder.py
index c576b55c1..b86be15a4 100644
--- a/agent/prompt_builder.py
+++ b/agent/prompt_builder.py
@@ -66,7 +66,8 @@ DEFAULT_AGENT_IDENTITY = (
     "range of tasks including answering questions, writing and editing code, "
     "analyzing information, creative work, and executing actions via your tools. "
     "You communicate clearly, admit uncertainty when appropriate, and prioritize "
-    "being genuinely useful over being verbose unless otherwise directed below."
+    "being genuinely useful over being verbose unless otherwise directed below. "
+    "Be targeted and efficient in your exploration and investigations."
 )
 
 MEMORY_GUIDANCE = (
diff --git a/batch_runner.py b/batch_runner.py
index 1bd6745b9..a4c402ffd 100644
--- a/batch_runner.py
+++ b/batch_runner.py
@@ -1155,7 +1155,7 @@ def main(
         providers_order (str): Comma-separated list of OpenRouter providers to try in order (e.g. "anthropic,openai,google")
         provider_sort (str): Sort providers by "price", "throughput", or "latency" (OpenRouter only)
         max_tokens (int): Maximum tokens for model responses (optional, uses model default if not set)
-        reasoning_effort (str): OpenRouter reasoning effort level: "xhigh", "high", "medium", "low", "minimal", "none" (default: "xhigh")
+        reasoning_effort (str): OpenRouter reasoning effort level: "xhigh", "high", "medium", "low", "minimal", "none" (default: "medium")
         reasoning_disabled (bool): Completely disable reasoning/thinking tokens (default: False)
         prefill_messages_file (str): Path to JSON file containing prefill messages (list of {role, content} dicts)
         max_samples (int): Only process the first N samples from the dataset (optional, processes all if not set)
@@ -1216,7 +1216,7 @@ def main(
     providers_order_list = [p.strip() for p in providers_order.split(",")] if providers_order else None
     
     # Build reasoning_config from CLI flags
-    # --reasoning_disabled takes priority, then --reasoning_effort, then default (xhigh)
+    # --reasoning_disabled takes priority, then --reasoning_effort, then default (medium)
     reasoning_config = None
     if reasoning_disabled:
         # Completely disable reasoning/thinking tokens
diff --git a/cli-config.yaml.example b/cli-config.yaml.example
index d8489d95b..f0d5a95bd 100644
--- a/cli-config.yaml.example
+++ b/cli-config.yaml.example
@@ -285,7 +285,7 @@ agent:
   # Reasoning effort level (OpenRouter and Nous Portal)
   # Controls how much "thinking" the model does before responding.
   # Options: "xhigh" (max), "high", "medium", "low", "minimal", "none" (disable)
-  reasoning_effort: "xhigh"
+  reasoning_effort: "medium"
   
   # Predefined personalities (use with /personality command)
   personalities:
diff --git a/cli.py b/cli.py
index 4d1941f81..7dd74b0b2 100755
--- a/cli.py
+++ b/cli.py
@@ -108,7 +108,7 @@ def _parse_reasoning_config(effort: str) -> dict | None:
     """Parse a reasoning effort level into an OpenRouter reasoning config dict.
     
     Valid levels: "xhigh", "high", "medium", "low", "minimal", "none".
-    Returns None to use the default (xhigh), or a config dict to override.
+    Returns None to use the default (medium), or a config dict to override.
     """
     if not effort or not effort.strip():
         return None
@@ -118,7 +118,7 @@ def _parse_reasoning_config(effort: str) -> dict | None:
     valid = ("xhigh", "high", "medium", "low", "minimal")
     if effort in valid:
         return {"enabled": True, "effort": effort}
-    logger.warning("Unknown reasoning_effort '%s', using default (xhigh)", effort)
+    logger.warning("Unknown reasoning_effort '%s', using default (medium)", effort)
     return None
 
 
diff --git a/gateway/run.py b/gateway/run.py
index 99fd2443f..3ed81379a 100644
--- a/gateway/run.py
+++ b/gateway/run.py
@@ -330,7 +330,7 @@ class GatewayRunner:
         
         Checks HERMES_REASONING_EFFORT env var first, then agent.reasoning_effort
         in config.yaml. Valid: "xhigh", "high", "medium", "low", "minimal", "none".
-        Returns None to use default (xhigh).
+        Returns None to use default (medium).
         """
         effort = os.getenv("HERMES_REASONING_EFFORT", "")
         if not effort:
@@ -351,7 +351,7 @@ class GatewayRunner:
         valid = ("xhigh", "high", "medium", "low", "minimal")
         if effort in valid:
             return {"enabled": True, "effort": effort}
-        logger.warning("Unknown reasoning_effort '%s', using default (xhigh)", effort)
+        logger.warning("Unknown reasoning_effort '%s', using default (medium)", effort)
         return None
 
     @staticmethod
diff --git a/run_agent.py b/run_agent.py
index 00c43657b..2fce80a9a 100644
--- a/run_agent.py
+++ b/run_agent.py
@@ -213,7 +213,7 @@ class AIAgent:
                 Provided by the platform layer (CLI or gateway). If None, the clarify tool returns an error.
             max_tokens (int): Maximum tokens for model responses (optional, uses model default if not set)
             reasoning_config (Dict): OpenRouter reasoning configuration override (e.g. {"effort": "none"} to disable thinking).
-                If None, defaults to {"enabled": True, "effort": "xhigh"} for OpenRouter. Set to disable/customize reasoning.
+                If None, defaults to {"enabled": True, "effort": "medium"} for OpenRouter. Set to disable/customize reasoning.
             prefill_messages (List[Dict]): Messages to prepend to conversation history as prefilled context.
                 Useful for injecting a few-shot example or priming the model's response style.
                 Example: [{"role": "user", "content": "Hi!"}, {"role": "assistant", "content": "Hello!"}]
@@ -287,7 +287,7 @@ class AIAgent:
         
         # Model response configuration
         self.max_tokens = max_tokens  # None = use model default
-        self.reasoning_config = reasoning_config  # None = use default (xhigh for OpenRouter)
+        self.reasoning_config = reasoning_config  # None = use default (medium for OpenRouter)
         self.prefill_messages = prefill_messages or []  # Prefilled conversation turns
         
         # Anthropic prompt caching: auto-enabled for Claude models via OpenRouter.
@@ -2157,8 +2157,8 @@ class AIAgent:
             if not instructions:
                 instructions = DEFAULT_AGENT_IDENTITY
 
-            # Resolve reasoning effort: config > default (xhigh)
-            reasoning_effort = "xhigh"
+            # Resolve reasoning effort: config > default (medium)
+            reasoning_effort = "medium"
             reasoning_enabled = True
             if self.reasoning_config and isinstance(self.reasoning_config, dict):
                 if self.reasoning_config.get("enabled") is False:
@@ -2224,7 +2224,7 @@ class AIAgent:
             else:
                 extra_body["reasoning"] = {
                     "enabled": True,
-                    "effort": "xhigh"
+                    "effort": "medium"
                 }
 
         # Nous Portal product attribution
@@ -2767,7 +2767,7 @@ class AIAgent:
                 else:
                     summary_extra_body["reasoning"] = {
                         "enabled": True,
-                        "effort": "xhigh"
+                        "effort": "medium"
                     }
             if _is_nous:
                 summary_extra_body["tags"] = ["product=hermes-agent"]
diff --git a/tests/test_provider_parity.py b/tests/test_provider_parity.py
index 00fc4dd9b..2ee313144 100644
--- a/tests/test_provider_parity.py
+++ b/tests/test_provider_parity.py
@@ -145,7 +145,7 @@ class TestBuildApiKwargsCodex:
         messages = [{"role": "user", "content": "hi"}]
         kwargs = agent._build_api_kwargs(messages)
         assert "reasoning" in kwargs
-        assert kwargs["reasoning"]["effort"] == "xhigh"
+        assert kwargs["reasoning"]["effort"] == "medium"
 
     def test_includes_encrypted_content_in_include(self, monkeypatch):
         agent = _make_agent(monkeypatch, "openai-codex", api_mode="codex_responses",
@@ -596,19 +596,19 @@ class TestCodexReasoningPreflight:
 # ── Reasoning effort consistency tests ───────────────────────────────────────
 
 class TestReasoningEffortDefaults:
-    """Verify reasoning effort defaults to xhigh across all provider paths."""
+    """Verify reasoning effort defaults to medium across all provider paths."""
 
-    def test_openrouter_default_xhigh(self, monkeypatch):
+    def test_openrouter_default_medium(self, monkeypatch):
         agent = _make_agent(monkeypatch, "openrouter")
         kwargs = agent._build_api_kwargs([{"role": "user", "content": "hi"}])
         reasoning = kwargs["extra_body"]["reasoning"]
-        assert reasoning["effort"] == "xhigh"
+        assert reasoning["effort"] == "medium"
 
-    def test_codex_default_xhigh(self, monkeypatch):
+    def test_codex_default_medium(self, monkeypatch):
         agent = _make_agent(monkeypatch, "openai-codex", api_mode="codex_responses",
                             base_url="https://chatgpt.com/backend-api/codex")
         kwargs = agent._build_api_kwargs([{"role": "user", "content": "hi"}])
-        assert kwargs["reasoning"]["effort"] == "xhigh"
+        assert kwargs["reasoning"]["effort"] == "medium"
 
     def test_codex_reasoning_disabled(self, monkeypatch):
         agent = _make_agent(monkeypatch, "openai-codex", api_mode="codex_responses",
diff --git a/tests/test_run_agent.py b/tests/test_run_agent.py
index ae7924d45..226b29a6d 100644
--- a/tests/test_run_agent.py
+++ b/tests/test_run_agent.py
@@ -498,12 +498,12 @@ class TestBuildApiKwargs:
         assert kwargs["extra_body"]["provider"]["only"] == ["Anthropic"]
 
     def test_reasoning_config_default_openrouter(self, agent):
-        """Default reasoning config for OpenRouter should be xhigh."""
+        """Default reasoning config for OpenRouter should be medium."""
         messages = [{"role": "user", "content": "hi"}]
         kwargs = agent._build_api_kwargs(messages)
         reasoning = kwargs["extra_body"]["reasoning"]
         assert reasoning["enabled"] is True
-        assert reasoning["effort"] == "xhigh"
+        assert reasoning["effort"] == "medium"
 
     def test_reasoning_config_custom(self, agent):
         agent.reasoning_config = {"enabled": False}
diff --git a/website/docs/user-guide/configuration.md b/website/docs/user-guide/configuration.md
index 6d6897794..33193619c 100644
--- a/website/docs/user-guide/configuration.md
+++ b/website/docs/user-guide/configuration.md
@@ -421,10 +421,10 @@ Control how much "thinking" the model does before responding:
 
 ```yaml
 agent:
-  reasoning_effort: ""   # empty = use model default. Options: xhigh (max), high, medium, low, minimal, none
+  reasoning_effort: ""   # empty = medium (default). Options: xhigh (max), high, medium, low, minimal, none
 ```
 
-When unset (default), the model's own default reasoning level is used. Setting a value overrides it — higher reasoning effort gives better results on complex tasks at the cost of more tokens and latency.
+When unset (default), reasoning effort defaults to "medium" — a balanced level that works well for most tasks. Setting a value overrides it — higher reasoning effort gives better results on complex tasks at the cost of more tokens and latency.
 
 ## TTS Configuration