diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/utils/repair-tool-call.util.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/utils/repair-tool-call.util.ts
index ddf71c0b63..5a4af61549 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/utils/repair-tool-call.util.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-agent/utils/repair-tool-call.util.ts
@@ -1,6 +1,17 @@
-import { type LanguageModel, NoSuchToolError, Output, generateText } from 'ai';
+import {
+  type LanguageModel,
+  type LanguageModelUsage,
+  NoSuchToolError,
+  Output,
+  type StepResult,
+  type ToolSet,
+  generateText,
+} from 'ai';
 import { type z } from 'zod';
 
+import { UsageOperationType } from 'src/engine/core-modules/usage/enums/usage-operation-type.enum';
+import { AiBillingService } from 'src/engine/metadata-modules/ai/ai-billing/services/ai-billing.service';
+import { extractCacheCreationTokensFromSteps } from 'src/engine/metadata-modules/ai/ai-billing/utils/extract-cache-creation-tokens.util';
 import { AI_TELEMETRY_CONFIG } from 'src/engine/metadata-modules/ai/ai-models/constants/ai-telemetry.const';
 
 type ToolCall = {
@@ -10,18 +21,28 @@ type ToolCall = {
   input: string;
 };
 
+type RepairToolCallBillingContext = {
+  aiBillingService: AiBillingService;
+  modelId: string;
+  workspaceId: string;
+  userWorkspaceId: string | null;
+  operationType: UsageOperationType;
+};
+
 export const repairToolCall = async ({
   toolCall,
   tools,
   inputSchema,
   error,
   model,
+  billingContext,
 }: {
   toolCall: ToolCall;
   tools: Record<string, unknown>;
   inputSchema: (toolCall: { toolName: string }) => unknown;
   error: Error;
   model: LanguageModel;
+  billingContext?: RepairToolCallBillingContext;
 }): Promise<ToolCall | null> => {
   // Don't attempt to fix invalid tool names
   if (NoSuchToolError.isInstance(error)) {
@@ -40,8 +61,11 @@ export const repairToolCall = async ({
     return null;
   }
 
+  let usage: LanguageModelUsage | undefined;
+  let steps: StepResult<ToolSet>[] | undefined;
+
   try {
-    const { output: repairedInput } = await generateText({
+    const result = await generateText({
       model,
       output: Output.object({ schema: schema as z.ZodTypeAny }),
       prompt: [
@@ -62,6 +86,11 @@ export const repairToolCall = async ({
       experimental_telemetry: AI_TELEMETRY_CONFIG,
     });
 
+    usage = result.usage;
+    steps = result.steps;
+
+    const repairedInput = result.output;
+
     if (repairedInput == null) {
       return null;
     }
@@ -75,5 +104,20 @@ export const repairToolCall = async ({
   } catch {
     // If repair fails, return null to let the error propagate
     return null;
+  } finally {
+    if (billingContext && usage) {
+      const cacheCreationTokens = steps
+        ? extractCacheCreationTokensFromSteps(steps)
+        : 0;
+
+      billingContext.aiBillingService.calculateAndBillUsage(
+        billingContext.modelId,
+        { usage, cacheCreationTokens },
+        billingContext.workspaceId,
+        billingContext.operationType,
+        null,
+        billingContext.userWorkspaceId,
+      );
+    }
   }
 };
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat.service.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat.service.ts
index 621970a50a..4f5fbfa7bb 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat.service.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-chat.service.ts
@@ -328,8 +328,11 @@ export class AgentChatService {
       return null;
     }
 
-    const title =
-      await this.titleGenerationService.generateThreadTitle(messageContent);
+    const title = await this.titleGenerationService.generateThreadTitle(
+      messageContent,
+      workspaceId,
+      thread.userWorkspaceId,
+    );
 
     await this.threadRepository.update(threadId, { title });
 
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-title-generation.service.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-title-generation.service.ts
index 8fed08f177..736a0c82f8 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-title-generation.service.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/agent-title-generation.service.ts
@@ -1,7 +1,15 @@
 import { Injectable, Logger } from '@nestjs/common';
 
-import { generateText } from 'ai';
+import {
+  type LanguageModelUsage,
+  type StepResult,
+  type ToolSet,
+  generateText,
+} from 'ai';
 
+import { UsageOperationType } from 'src/engine/core-modules/usage/enums/usage-operation-type.enum';
+import { AiBillingService } from 'src/engine/metadata-modules/ai/ai-billing/services/ai-billing.service';
+import { extractCacheCreationTokensFromSteps } from 'src/engine/metadata-modules/ai/ai-billing/utils/extract-cache-creation-tokens.util';
 import { AI_TELEMETRY_CONFIG } from 'src/engine/metadata-modules/ai/ai-models/constants/ai-telemetry.const';
 import { AiModelRegistryService } from 'src/engine/metadata-modules/ai/ai-models/services/ai-model-registry.service';
 
@@ -11,29 +19,55 @@ export class AgentTitleGenerationService {
 
   constructor(
     private readonly aiModelRegistryService: AiModelRegistryService,
+    private readonly aiBillingService: AiBillingService,
   ) {}
 
-  async generateThreadTitle(messageContent: string): Promise<string> {
+  async generateThreadTitle(
+    messageContent: string,
+    workspaceId: string,
+    userWorkspaceId: string | null,
+  ): Promise<string> {
+    const defaultModel = this.aiModelRegistryService.getDefaultSpeedModel();
+
+    if (!defaultModel) {
+      this.logger.warn('No default AI model available for title generation');
+
+      return this.generateFallbackTitle(messageContent);
+    }
+
+    let usage: LanguageModelUsage | undefined;
+    let steps: StepResult<ToolSet>[] | undefined;
+
     try {
-      const defaultModel = this.aiModelRegistryService.getDefaultSpeedModel();
-
-      if (!defaultModel) {
-        this.logger.warn('No default AI model available for title generation');
-
-        return this.generateFallbackTitle(messageContent);
-      }
-
       const result = await generateText({
         model: defaultModel.model,
         prompt: `Generate a concise, descriptive title (maximum 60 characters) for a chat thread based on the following message. The title should capture the main topic or purpose of the conversation. Return only the title, nothing else. Message: "${messageContent}"`,
         experimental_telemetry: AI_TELEMETRY_CONFIG,
       });
 
+      usage = result.usage;
+      steps = result.steps;
+
       return this.cleanTitle(result.text);
     } catch (error) {
       this.logger.error('Failed to generate title with AI:', error);
 
       return this.generateFallbackTitle(messageContent);
+    } finally {
+      if (usage) {
+        const cacheCreationTokens = steps
+          ? extractCacheCreationTokensFromSteps(steps)
+          : 0;
+
+        this.aiBillingService.calculateAndBillUsage(
+          defaultModel.modelId,
+          { usage, cacheCreationTokens },
+          workspaceId,
+          UsageOperationType.AI_CHAT_TOKEN,
+          null,
+          userWorkspaceId,
+        );
+      }
     }
   }
 
diff --git a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts
index 8905a72f0e..5cabaac7f5 100644
--- a/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts
+++ b/packages/twenty-server/src/engine/metadata-modules/ai/ai-chat/services/chat-execution.service.ts
@@ -348,6 +348,13 @@ export class ChatExecutionService {
           inputSchema,
           error,
           model: registeredModel.model,
+          billingContext: {
+            aiBillingService: this.aiBillingService,
+            modelId: registeredModel.modelId,
+            workspaceId: workspace.id,
+            userWorkspaceId,
+            operationType: UsageOperationType.AI_CHAT_TOKEN,
+          },
         });
       },
     });