diff --git a/ts/api/handlers/chat.ts b/ts/api/handlers/chat.ts
index ede95a9..4d566d9 100644
--- a/ts/api/handlers/chat.ts
+++ b/ts/api/handlers/chat.ts
@@ -6,6 +6,7 @@ import * as http from 'node:http';
 import type { IApiError, IChatCompletionRequest } from '../../interfaces/api.ts';
 import { ClusterCoordinator } from '../../cluster/coordinator.ts';
 import { ContainerManager } from '../../containers/container-manager.ts';
+import { API_SERVER } from '../../constants.ts';
 import { logger } from '../../logger.ts';
 import { ModelRegistry } from '../../models/registry.ts';
 import { ModelLoader } from '../../models/loader.ts';
@@ -158,11 +159,14 @@ export class ChatHandler {
     targetEndpoint: string,
     body: IChatCompletionRequest,
   ): Promise<void> {
+    const controller = new AbortController();
+    const timeout = setTimeout(() => controller.abort(), API_SERVER.REQUEST_TIMEOUT_MS);
     const response = await fetch(`${targetEndpoint}/v1/chat/completions`, {
       method: 'POST',
       headers: this.buildForwardHeaders(req),
       body: JSON.stringify(body),
-    });
+      signal: controller.signal,
+    }).finally(() => clearTimeout(timeout));
 
     if (body.stream) {
       res.writeHead(response.status, {
diff --git a/ts/api/handlers/embeddings.ts b/ts/api/handlers/embeddings.ts
index 9873c50..5c90a76 100644
--- a/ts/api/handlers/embeddings.ts
+++ b/ts/api/handlers/embeddings.ts
@@ -11,6 +11,7 @@ import type {
 } from '../../interfaces/api.ts';
 import { ClusterCoordinator } from '../../cluster/coordinator.ts';
 import { ContainerManager } from '../../containers/container-manager.ts';
+import { API_SERVER } from '../../constants.ts';
 import { logger } from '../../logger.ts';
 import { ModelRegistry } from '../../models/registry.ts';
 
@@ -80,7 +81,7 @@ export class EmbeddingsHandler {
         return;
       }
 
-      const response = await fetch(`${ensured.location.endpoint}/v1/embeddings`, {
+      const response = await this.fetchWithTimeout(`${ensured.location.endpoint}/v1/embeddings`, {
         method: 'POST',
         headers: this.buildForwardHeaders(req),
         body: JSON.stringify(requestBody),
@@ -159,7 +160,7 @@ export class EmbeddingsHandler {
     model: string,
     input: string,
   ): Promise<{ vector: number[]; tokenCount: number }> {
-    const response = await fetch(`${endpoint}/v1/embeddings`, {
+    const response = await this.fetchWithTimeout(`${endpoint}/v1/embeddings`, {
       method: 'POST',
       headers: { 'Content-Type': 'application/json' },
       body: JSON.stringify({ model, input }),
@@ -181,7 +182,7 @@ export class EmbeddingsHandler {
     _model: string,
     input: string,
   ): Promise<{ vector: number[]; tokenCount: number }> {
-    const response = await fetch(`${endpoint}/embed`, {
+    const response = await this.fetchWithTimeout(`${endpoint}/embed`, {
       method: 'POST',
       headers: { 'Content-Type': 'application/json' },
       body: JSON.stringify({ inputs: input }),
@@ -214,6 +215,20 @@ export class EmbeddingsHandler {
     return headers;
   }
 
+  private async fetchWithTimeout(url: string, init: RequestInit): Promise<Response> {
+    const controller = new AbortController();
+    const timeout = setTimeout(() => controller.abort(), API_SERVER.REQUEST_TIMEOUT_MS);
+
+    try {
+      return await fetch(url, {
+        ...init,
+        signal: controller.signal,
+      });
+    } finally {
+      clearTimeout(timeout);
+    }
+  }
+
   private sendError(
     res: http.ServerResponse,
     statusCode: number,
diff --git a/ts/containers/base-container.ts b/ts/containers/base-container.ts
index d3e68d3..eb676a1 100644
--- a/ts/containers/base-container.ts
+++ b/ts/containers/base-container.ts
@@ -11,6 +11,7 @@ import type {
   TContainerType,
 } from '../interfaces/container.ts';
 import type { IChatCompletionRequest, IChatCompletionResponse } from '../interfaces/api.ts';
+import { API_SERVER } from '../constants.ts';
 import { ContainerRuntime } from '../docker/container-runtime.ts';
 import { logger } from '../logger.ts';
 
@@ -165,7 +166,7 @@ export abstract class BaseContainer {
     const url = `${endpoint}${path}`;
 
     const controller = new AbortController();
-    const timeout = options.timeout || 30000;
+    const timeout = options.timeout || API_SERVER.REQUEST_TIMEOUT_MS;
     const timeoutId = setTimeout(() => controller.abort(), timeout);
 
     try {
diff --git a/ts/models/registry.ts b/ts/models/registry.ts
index 9e68629..46c7e4e 100644
--- a/ts/models/registry.ts
+++ b/ts/models/registry.ts
@@ -4,7 +4,7 @@
 
 import * as fs from 'node:fs/promises';
 import type { IModelCatalog, IModelCatalogEntry } from '../interfaces/catalog.ts';
-import { MODEL_REGISTRY, TIMING } from '../constants.ts';
+import { API_SERVER, MODEL_REGISTRY, TIMING } from '../constants.ts';
 import { logger } from '../logger.ts';
 
 export class ModelRegistry {
@@ -167,7 +167,7 @@ export class ModelRegistry {
     }
 
     const controller = new AbortController();
-    const timeout = setTimeout(() => controller.abort(), 30000);
+    const timeout = setTimeout(() => controller.abort(), API_SERVER.REQUEST_TIMEOUT_MS);
 
     try {
       const response = await fetch(source, {