diff --git a/agent/context_compressor.py b/agent/context_compressor.py
index a0ca0c99..b2dff9c8 100644
--- a/agent/context_compressor.py
+++ b/agent/context_compressor.py
@@ -28,7 +28,7 @@ class ContextCompressor:
     def __init__(
         self,
         model: str,
-        threshold_percent: float = 0.85,
+        threshold_percent: float = 0.50,
         protect_first_n: int = 3,
         protect_last_n: int = 4,
         summary_target_tokens: int = 2500,
diff --git a/cli.py b/cli.py
index 1e418177..04794230 100755
--- a/cli.py
+++ b/cli.py
@@ -175,7 +175,7 @@ def load_cli_config() -> Dict[str, Any]:
         },
         "compression": {
             "enabled": True,      # Auto-compress when approaching context limit
-            "threshold": 0.85,    # Compress at 85% of model's context limit
+            "threshold": 0.50,    # Compress at 50% of model's context limit
             "summary_model": "google/gemini-3-flash-preview",  # Fast/cheap model for summaries
         },
         "agent": {
diff --git a/gateway/run.py b/gateway/run.py
index 8c068559..07490b77 100644
--- a/gateway/run.py
+++ b/gateway/run.py
@@ -1057,7 +1057,7 @@ class GatewayRunner:
             # Read model + compression config from config.yaml — same
             # source of truth the agent itself uses.
             _hyg_model = "anthropic/claude-sonnet-4.6"
-            _hyg_threshold_pct = 0.85
+            _hyg_threshold_pct = 0.50
             _hyg_compression_enabled = True
             try:
                 _hyg_cfg_path = _hermes_home / "config.yaml"
diff --git a/hermes_cli/config.py b/hermes_cli/config.py
index 39501e93..c05ebd5a 100644
--- a/hermes_cli/config.py
+++ b/hermes_cli/config.py
@@ -121,7 +121,7 @@ DEFAULT_CONFIG = {
     
     "compression": {
         "enabled": True,
-        "threshold": 0.85,
+        "threshold": 0.50,
         "summary_model": "google/gemini-3-flash-preview",
         "summary_provider": "auto",
     },
@@ -1119,7 +1119,7 @@ def show_config():
     enabled = compression.get('enabled', True)
     print(f"  Enabled:      {'yes' if enabled else 'no'}")
     if enabled:
-        print(f"  Threshold:    {compression.get('threshold', 0.85) * 100:.0f}%")
+        print(f"  Threshold:    {compression.get('threshold', 0.50) * 100:.0f}%")
         print(f"  Model:        {compression.get('summary_model', 'google/gemini-3-flash-preview')}")
         comp_provider = compression.get('summary_provider', 'auto')
         if comp_provider != 'auto':
diff --git a/run_agent.py b/run_agent.py
index a07cc472..2f082571 100644
--- a/run_agent.py
+++ b/run_agent.py
@@ -664,7 +664,7 @@ class AIAgent:
         # Initialize context compressor for automatic context management
         # Compresses conversation when approaching model's context limit
         # Configuration via config.yaml (compression section) or environment variables
-        compression_threshold = float(os.getenv("CONTEXT_COMPRESSION_THRESHOLD", "0.85"))
+        compression_threshold = float(os.getenv("CONTEXT_COMPRESSION_THRESHOLD", "0.50"))
         compression_enabled = os.getenv("CONTEXT_COMPRESSION_ENABLED", "true").lower() in ("true", "1", "yes")
         compression_summary_model = os.getenv("CONTEXT_COMPRESSION_MODEL") or None