diff --git a/docs/tools/thinking.md b/docs/tools/thinking.md
index 22cced0a5ba..91c3359f552 100644
--- a/docs/tools/thinking.md
+++ b/docs/tools/thinking.md
@@ -15,13 +15,14 @@ title: "Thinking Levels"
   - low → “think hard”
   - medium → “think harder”
   - high → “ultrathink” (max budget)
-  - xhigh → “ultrathink+” (GPT-5.2 + Codex models and Anthropic Claude Opus 4.7)
-  - adaptive → provider-managed adaptive reasoning budget (supported for Anthropic Claude 4.6 and Opus 4.7)
+  - xhigh → “ultrathink+” (GPT-5.2 + Codex models and Anthropic Claude Opus 4.7 effort)
+  - adaptive → provider-managed adaptive thinking (supported for Anthropic Claude 4.6 and Opus 4.7)
   - `x-high`, `x_high`, `extra-high`, `extra high`, and `extra_high` map to `xhigh`.
   - `highest`, `max` map to `high`.
 - Provider notes:
-  - Anthropic Claude 4.6 and Opus 4.7 models default to `adaptive` when no explicit thinking level is set.
-  - Anthropic Claude Opus 4.7 maps `/think xhigh` to `output_config.effort: "xhigh"`.
+  - Anthropic Claude 4.6 models default to `adaptive` when no explicit thinking level is set.
+  - Anthropic Claude Opus 4.7 does not default to adaptive thinking. Its API effort default remains provider-owned unless you explicitly set a thinking level.
+  - Anthropic Claude Opus 4.7 maps `/think xhigh` to adaptive thinking plus `output_config.effort: "xhigh"`, because `/think` is a thinking directive and `xhigh` is the Opus 4.7 effort setting.
   - MiniMax (`minimax/*`) on the Anthropic-compatible streaming path defaults to `thinking: { type: "disabled" }` unless you explicitly set thinking in model params or request params. This avoids leaked `reasoning_content` deltas from MiniMax's non-native Anthropic stream format.
   - Z.AI (`zai/*`) only supports binary thinking (`on`/`off`). Any non-`off` level is treated as `on` (mapped to `low`).
   - Moonshot (`moonshot/*`) maps `/think off` to `thinking: { type: "disabled" }` and any non-`off` level to `thinking: { type: "enabled" }`. When thinking is enabled, Moonshot only accepts `tool_choice` `auto|none`; OpenClaw normalizes incompatible values to `auto`.
@@ -32,7 +33,7 @@ title: "Thinking Levels"
 2. Session override (set by sending a directive-only message).
 3. Per-agent default (`agents.list[].thinkingDefault` in config).
 4. Global default (`agents.defaults.thinkingDefault` in config).
-5. Fallback: `adaptive` for Anthropic Claude 4.6 and Opus 4.7 models, `low` for other reasoning-capable models, `off` otherwise.
+5. Fallback: `adaptive` for Anthropic Claude 4.6 models, `off` for Anthropic Claude Opus 4.7 unless explicitly configured, `low` for other reasoning-capable models, `off` otherwise.
 
 ## Setting a session default
 
@@ -105,7 +106,7 @@ title: "Thinking Levels"
 
 - The web chat thinking selector mirrors the session's stored level from the inbound session store/config when the page loads.
 - Picking another level writes the session override immediately via `sessions.patch`; it does not wait for the next send and it is not a one-shot `thinkingOnce` override.
-- The first option is always `Default (<resolved level>)`, where the resolved default comes from the active session model: `adaptive` for Claude 4.6 and Opus 4.7 on Anthropic, `low` for other reasoning-capable models, `off` otherwise.
+- The first option is always `Default (<resolved level>)`, where the resolved default comes from the active session model: `adaptive` for Claude 4.6 on Anthropic, `off` for Anthropic Claude Opus 4.7 unless configured, `low` for other reasoning-capable models, `off` otherwise.
 - The picker stays provider-aware:
   - most providers show `off | minimal | low | medium | high | adaptive`
   - Anthropic Claude Opus 4.7 shows `off | minimal | low | medium | high | xhigh | adaptive`
diff --git a/extensions/anthropic/index.test.ts b/extensions/anthropic/index.test.ts
index c8ec785be1f..daaa429d0de 100644
--- a/extensions/anthropic/index.test.ts
+++ b/extensions/anthropic/index.test.ts
@@ -199,6 +199,12 @@ describe("anthropic provider replay hooks", () => {
         provider: "anthropic",
         modelId: "claude-opus-4-7",
       } as never),
+    ).toBe("off");
+    expect(
+      provider.resolveDefaultThinkingLevel?.({
+        provider: "anthropic",
+        modelId: "claude-opus-4-6",
+      } as never),
     ).toBe("adaptive");
     expect(
       provider.supportsXHighThinking?.({
diff --git a/extensions/anthropic/register.runtime.ts b/extensions/anthropic/register.runtime.ts
index 79e13d96d66..510ffaf2883 100644
--- a/extensions/anthropic/register.runtime.ts
+++ b/extensions/anthropic/register.runtime.ts
@@ -260,7 +260,6 @@ function resolveAnthropicForwardCompatModel(
 function shouldUseAnthropicAdaptiveThinkingDefault(modelId: string): boolean {
   const lowerModelId = normalizeLowercaseStringOrEmpty(modelId);
   return (
-    isAnthropicOpus47Model(lowerModelId) ||
     lowerModelId.startsWith(ANTHROPIC_OPUS_46_MODEL_ID) ||
     lowerModelId.startsWith(ANTHROPIC_OPUS_46_DOT_MODEL_ID) ||
     lowerModelId.startsWith(ANTHROPIC_SONNET_46_MODEL_ID) ||
@@ -491,9 +490,11 @@ export function registerAnthropicPlugin(api: OpenClawPluginApi): void {
     supportsXHighThinking: ({ modelId }) => isAnthropicOpus47Model(modelId),
     wrapStreamFn: wrapAnthropicProviderStream,
     resolveDefaultThinkingLevel: ({ modelId }) =>
-      matchesAnthropicModernModel(modelId) && shouldUseAnthropicAdaptiveThinkingDefault(modelId)
-        ? "adaptive"
-        : undefined,
+      isAnthropicOpus47Model(modelId)
+        ? "off"
+        : matchesAnthropicModernModel(modelId) && shouldUseAnthropicAdaptiveThinkingDefault(modelId)
+          ? "adaptive"
+          : undefined,
     resolveUsageAuth: async (ctx) => await ctx.resolveOAuthToken(),
     fetchUsageSnapshot: async (ctx) =>
       await fetchClaudeUsage(ctx.token, ctx.timeoutMs, ctx.fetchFn),
diff --git a/src/agents/model-selection.test.ts b/src/agents/model-selection.test.ts
index 51df92af395..e7e0d350434 100644
--- a/src/agents/model-selection.test.ts
+++ b/src/agents/model-selection.test.ts
@@ -1176,7 +1176,7 @@ describe("model-selection", () => {
       expect(resolveAnthropicOpusThinking(cfg)).toBe("adaptive");
     });
 
-    it("uses adaptive fallback for explicitly configured Anthropic Opus 4.7", () => {
+    it("keeps thinking off by default for explicitly configured Anthropic Opus 4.7", () => {
       const cfg = {
         agents: {
           defaults: {
@@ -1185,7 +1185,7 @@ describe("model-selection", () => {
         },
       } as OpenClawConfig;
 
-      expect(resolveAnthropicOpus47Thinking(cfg)).toBe("adaptive");
+      expect(resolveAnthropicOpus47Thinking(cfg)).toBe("off");
     });
 
     it("falls back to low when no provider thinking hook is active", () => {
diff --git a/src/agents/model-thinking-default.ts b/src/agents/model-thinking-default.ts
index 7ea7e0179d7..35db5ca1935 100644
--- a/src/agents/model-thinking-default.ts
+++ b/src/agents/model-thinking-default.ts
@@ -54,13 +54,18 @@ export function resolveThinkingDefault(params: {
   if (configured) {
     return configured;
   }
+  if (
+    normalizedProvider === "anthropic" &&
+    (normalizedModel.startsWith("claude-opus-4-7") || normalizedModel.startsWith("claude-opus-4.7"))
+  ) {
+    return "off";
+  }
   if (
     normalizedProvider === "anthropic" &&
     explicitModelConfigured &&
     typeof catalogCandidate?.name === "string" &&
-    /4\.[67]\b/.test(catalogCandidate.name) &&
-    (normalizedModel.startsWith("claude-opus-4-7") ||
-      normalizedModel.startsWith("claude-opus-4-6") ||
+    /4\.6\b/.test(catalogCandidate.name) &&
+    (normalizedModel.startsWith("claude-opus-4-6") ||
       normalizedModel.startsWith("claude-sonnet-4-6"))
   ) {
     return "adaptive";