diff --git a/README.md b/README.md
index 6b803ca..987b398 100644
--- a/README.md
+++ b/README.md
@@ -48,18 +48,9 @@ service.
 - `LLM_MODEL`, default `qwen2.5-14b`
 - `LLM_TIMEOUT`, default `5m`
 - `WHISPERX_URL`, WhisperX endpoint for transcription jobs
-- `OPENCLAW_URL`, optional OpenClaw gateway URL if we route through OpenClaw
-  instead of direct vLLM
 
 ## Next integration step
 
 `telephony` should first mirror low-risk analysis jobs into this service while
 continuing local processing. Remote execution can then be enabled by feature
 flag per task type.
-
-## OpenClaw note
-
-Current Portal services call the local AI server directly: vLLM for LLM tasks
-and WhisperX for transcription. OpenClaw is not required for the current
-`ai-service` queue deployment. It becomes useful if we want centralized model
-routing, provider fallback, request policy and cross-model gateway behavior.
diff --git a/internal/config/config.go b/internal/config/config.go
index 37cb844..db26a42 100644
--- a/internal/config/config.go
+++ b/internal/config/config.go
@@ -17,7 +17,6 @@ type Config struct {
 	LLMModel    string
 	LLMTimeout  time.Duration
 	WhisperXURL string
-	OpenClawURL string
 }
 
 func Load() Config {
@@ -32,7 +31,6 @@ func Load() Config {
 		LLMModel:    envString("LLM_MODEL", "qwen2.5-14b"),
 		LLMTimeout:  envDuration("LLM_TIMEOUT", 5*time.Minute),
 		WhisperXURL: envString("WHISPERX_URL", ""),
-		OpenClawURL: envString("OPENCLAW_URL", ""),
 	}
 }
 
diff --git a/internal/httpapi/providers.go b/internal/httpapi/providers.go
index 7ce3301..be14a38 100644
--- a/internal/httpapi/providers.go
+++ b/internal/httpapi/providers.go
@@ -34,7 +34,6 @@ func (s *Server) handleProviderStatus(w http.ResponseWriter, r *http.Request) {
 		Providers: []providerStatus{
 			s.checkLLM(ctx),
 			s.checkWhisperX(ctx),
-			s.checkOpenClaw(ctx),
 		},
 	}
 	writeJSON(w, http.StatusOK, resp)
@@ -102,58 +101,6 @@ func (s *Server) checkWhisperX(ctx context.Context) providerStatus {
 	return st
 }
 
-func (s *Server) checkOpenClaw(ctx context.Context) providerStatus {
-	baseURL := strings.TrimRight(strings.TrimSpace(s.cfg.OpenClawURL), "/")
-	st := providerStatus{Name: "openclaw", Configured: baseURL != "", URL: baseURL}
-	if !st.Configured {
-		return st
-	}
-	start := time.Now()
-	req, err := http.NewRequestWithContext(ctx, http.MethodGet, baseURL+"/healthz", nil)
-	if err != nil {
-		st.Error = err.Error()
-		return st
-	}
-	res, err := http.DefaultClient.Do(req)
-	st.LatencyMS = time.Since(start).Milliseconds()
-	if err != nil {
-		st.Error = err.Error()
-		return st
-	}
-	defer res.Body.Close()
-	if res.StatusCode == http.StatusNotFound {
-		return checkOpenClawHealth(ctx, baseURL, start)
-	}
-	if res.StatusCode >= 300 {
-		st.Error = fmt.Sprintf("http %d: %s", res.StatusCode, readSmallBody(res.Body))
-		return st
-	}
-	st.OK = true
-	return st
-}
-
-func checkOpenClawHealth(ctx context.Context, baseURL string, start time.Time) providerStatus {
-	st := providerStatus{Name: "openclaw", Configured: true, URL: baseURL}
-	req, err := http.NewRequestWithContext(ctx, http.MethodGet, baseURL+"/health", nil)
-	if err != nil {
-		st.Error = err.Error()
-		return st
-	}
-	res, err := http.DefaultClient.Do(req)
-	st.LatencyMS = time.Since(start).Milliseconds()
-	if err != nil {
-		st.Error = err.Error()
-		return st
-	}
-	defer res.Body.Close()
-	if res.StatusCode >= 300 {
-		st.Error = fmt.Sprintf("http %d: %s", res.StatusCode, readSmallBody(res.Body))
-		return st
-	}
-	st.OK = true
-	return st
-}
-
 func readSmallBody(r io.Reader) string {
 	body, err := io.ReadAll(io.LimitReader(r, 512))
 	if err != nil {
diff --git a/k8s/configmap.yaml b/k8s/configmap.yaml
index c4b9285..e7250de 100644
--- a/k8s/configmap.yaml
+++ b/k8s/configmap.yaml
@@ -7,10 +7,8 @@ data:
   HTTP_HOST: "0.0.0.0"
   HTTP_PORT: "8080"
   MIGRATE_ON_START: "true"
-  # Default direct AI endpoints. OpenClaw can replace LLM_BASE_URL later when
-  # we decide to route model traffic through a gateway instead of direct vLLM.
+  # Direct AI endpoints on the local AI server.
   LLM_BASE_URL: "http://10.2.3.5:8002"
   LLM_MODEL: "qwen2.5-14b"
   LLM_TIMEOUT: "5m"
   WHISPERX_URL: "http://10.2.3.5:8001"
-  OPENCLAW_URL: ""