diff --git a/conductor/tracks/context_token_viz_20260301/plan.md b/conductor/tracks/context_token_viz_20260301/plan.md
index 0487076..fb037c1 100644
--- a/conductor/tracks/context_token_viz_20260301/plan.md
+++ b/conductor/tracks/context_token_viz_20260301/plan.md
@@ -11,13 +11,13 @@ Architecture reference: [docs/guide_architecture.md](../../docs/guide_architectu
 
 ## Phase 2: Trimming Preview & Cache Status
 
-- [x] Task 2.1: When `stats.get('would_trim')` is True, render a warning: `imgui.text_colored(ImVec4(1,0.3,0,1), "WARNING: Next call will trim history")`. Below it, show `f"Trimmable turns: {stats['trimmable_turns']}"`. If `stats` contains per-message breakdown, render the first 3 trimmable messages with their role and token count in a compact list.
-- [x] Task 2.2: Add Gemini cache status display. Read `ai_client._gemini_cache` (check `is not None`), `ai_client._gemini_cache_created_at`, and `ai_client._GEMINI_CACHE_TTL`. If cache exists, show: `"Gemini Cache: ACTIVE | Age: {age_seconds}s / {ttl}s | Renews at: {ttl * 0.9:.0f}s"`. If not, show `"Gemini Cache: INACTIVE"`. Guard with `if ai_client._provider == "gemini":`.
-- [x] Task 2.3: Add Anthropic cache hint. When provider is `"anthropic"`, show: `"Anthropic: 4-breakpoint ephemeral caching (auto-managed)"` with the number of history turns and whether the latest response used cache reads (check last comms log entry for `cache_read_input_tokens`).
-- [~] Task 2.4: Write tests for trimming warning visibility and cache status display.
+- [x] Task 2.1: When `stats.get('would_trim')` is True. 7b5d9b1, render a warning: `imgui.text_colored(ImVec4(1,0.3,0,1), "WARNING: Next call will trim history")`. Below it, show `f"Trimmable turns: {stats['trimmable_turns']}"`. If `stats` contains per-message breakdown, render the first 3 trimmable messages with their role and token count in a compact list.
+- [x] Task 2.2: Add Gemini cache status display. 7b5d9b1 Read `ai_client._gemini_cache` (check `is not None`), `ai_client._gemini_cache_created_at`, and `ai_client._GEMINI_CACHE_TTL`. If cache exists, show: `"Gemini Cache: ACTIVE | Age: {age_seconds}s / {ttl}s | Renews at: {ttl * 0.9:.0f}s"`. If not, show `"Gemini Cache: INACTIVE"`. Guard with `if ai_client._provider == "gemini":`.
+- [x] Task 2.3: Add Anthropic cache hint. 7b5d9b1 When provider is `"anthropic"`, show: `"Anthropic: 4-breakpoint ephemeral caching (auto-managed)"` with the number of history turns and whether the latest response used cache reads (check last comms log entry for `cache_read_input_tokens`).
+- [x] Task 2.4: Write tests for trimming warning visibility and cache status display. 7b5d9b1
 
 ## Phase 3: Auto-Refresh & Integration
 
-- [ ] Task 3.1: Hook `_token_stats` refresh into three trigger points: (a) after `_do_generate()` completes — cache `stable_md` and call `get_history_bleed_stats`; (b) after provider/model switch in `current_provider.setter` and `current_model.setter` — clear and re-fetch; (c) after each `handle_ai_response` in `_process_pending_gui_tasks` — refresh stats since history grew. For (c), use a flag `self._token_stats_dirty = True` and refresh in the next frame's render call to avoid calling the stats function too frequently.
-- [ ] Task 3.2: Add the token budget panel to the Hook API. Extend `/api/gui/mma_status` (or add a new `/api/gui/token_stats` endpoint) to expose `_token_stats` for simulation verification. This allows tests to assert on token utilization levels.
+- [x] Task 3.1: Hook `_token_stats` refresh into three trigger points: (a) after `_do_generate()` completes — cache `stable_md` and call `get_history_bleed_stats`; (b) after provider/model switch in `current_provider.setter` and `current_model.setter` — clear and re-fetch; (c) after each `handle_ai_response` in `_process_pending_gui_tasks` — refresh stats since history grew. For (c), use a flag `self._token_stats_dirty = True` and refresh in the next frame's render call to avoid calling the stats function too frequently.
+- [~] Task 3.2: Add the token budget panel to the Hook API. Extend `/api/gui/mma_status` (or add a new `/api/gui/token_stats` endpoint) to expose `_token_stats` for simulation verification. This allows tests to assert on token utilization levels.
 - [ ] Task 3.3: Conductor - User Manual Verification 'Phase 3: Auto-Refresh & Integration' (Protocol in workflow.md)
diff --git a/gui_2.py b/gui_2.py
index 0517069..35fce3c 100644
--- a/gui_2.py
+++ b/gui_2.py
@@ -298,6 +298,7 @@ class App:
   self._gemini_cache_text = ""
   self._last_stable_md: str = ''
   self._token_stats: dict = {}
+  self._token_stats_dirty: bool = False
   self.ui_disc_truncate_pairs: int = 2
   self.ui_auto_scroll_comms = True
   self.ui_auto_scroll_tool_calls = True
@@ -360,7 +361,10 @@ class App:
     if hasattr(self, 'hook_server'):
      self.hook_server.start()
    self.available_models = []
+   self.available_models = []
    self._fetch_models(value)
+   self._token_stats = {}
+   self._token_stats_dirty = True
 
  @property
  def current_model(self) -> str:
@@ -372,6 +376,8 @@ class App:
    self._current_model = value
    ai_client.reset_session()
    ai_client.set_provider(self.current_provider, value)
+   self._token_stats = {}
+   self._token_stats_dirty = True
 
  def _init_ai_and_hooks(self) -> None:
   ai_client.set_provider(self.current_provider, self.current_model)
@@ -607,6 +613,12 @@ class App:
   async def stream(req: GenerateRequest) -> Any:
    """Placeholder for streaming AI generation responses (Not yet implemented)."""
    raise HTTPException(status_code=501, detail="Streaming endpoint (/api/v1/stream) is not yet supported in this version.")      
+
+  @api.get("/api/gui/token_stats", dependencies=[Depends(get_api_key)])
+  def token_stats() -> dict[str, Any]:
+   """Returns current token budget stats for simulation/test verification."""
+   return dict(self._token_stats)
+
   return api
   # ---------------------------------------------------------------- project loading
 
@@ -935,6 +947,8 @@ class App:
        self.ai_response = text
       self.ai_status = payload.get("status", "done")
      self._trigger_blink = True
+     if not stream_id:
+      self._token_stats_dirty = True
      if self.ui_auto_add_history and not stream_id:
       role = payload.get("role", "AI")
       with self._pending_history_adds_lock:
@@ -2732,6 +2746,9 @@ class App:
    imgui.text_colored(C_SUB, self._gemini_cache_text)
 
  def _render_token_budget_panel(self) -> None:
+  if self._token_stats_dirty:
+   self._token_stats_dirty = False
+   self._refresh_api_metrics({}, md_content=self._last_stable_md or None)
   stats = self._token_stats
   if not stats:
    imgui.text_disabled("Token stats unavailable")