diff --git a/speech/README.md b/speech/README.md
index 19ec303cda..c316d64ec3 100644
--- a/speech/README.md
+++ b/speech/README.md
@@ -32,7 +32,7 @@ recognition technologies into developer applications.
 
 View the [documentation][recognize_docs] or the [source code][recognize_code].
 
-__Usage:__ `node recognize --help`
+__Usage:__ `node recognize.js --help`
 
 ```
 Commands:
@@ -42,15 +42,13 @@ Commands:
   listen             Detects speech in a microphone input stream.
 
 Options:
-  --help  Show help                                                                                            [boolean]
+  --help  Show help                                                    [boolean]
 
 Examples:
-  node recognize sync ./resources/audio.raw    Detects speech in "./resources/audio.raw".
-  node recognize async ./resources/audio.raw   Creates a job to detect speech in "./resources/audio.raw", and waits for
-                                               the job to complete.
-  node recognize stream ./resources/audio.raw  Detects speech in "./resources/audio.raw" by streaming it to the Speech
-                                               API.
-  node recognize listen ./resources/audio.raw  Detects speech in a microphone input stream.
+  node recognize.js sync ./resources/audio.raw
+  node recognize.js async ./resources/audio.raw
+  node recognize.js stream ./resources/audio.raw
+  node recognize.js listen
 
 For more information, see https://cloud.google.com/speech/docs
 ```
diff --git a/speech/package.json b/speech/package.json
index 63b3f55254..4ce70045cb 100644
--- a/speech/package.json
+++ b/speech/package.json
@@ -5,17 +5,13 @@
   "license": "Apache Version 2.0",
   "author": "Google Inc.",
   "scripts": {
-    "test": "mocha -R spec -t 10000 --require intelli-espower-loader ../test/_setup.js test/*.test.js",
-    "system-test": "mocha -R spec -t 10000 --require intelli-espower-loader ../system-test/_setup.js system-test/*.test.js"
+    "test": "cd ..; npm run st -- speech/system-test/*.test.js"
   },
   "dependencies": {
     "@google-cloud/speech": "^0.4.0",
     "node-record-lpcm16": "^0.1.4",
     "yargs": "^6.4.0"
   },
-  "devDependencies": {
-    "mocha": "^3.1.2"
-  },
   "engines": {
     "node": ">=4.3.2"
   }
diff --git a/speech/quickstart.js b/speech/quickstart.js
index fa28527ca3..bc2cde2bd4 100644
--- a/speech/quickstart.js
+++ b/speech/quickstart.js
@@ -37,12 +37,9 @@ const options = {
 };
 
 // Detects speech in the audio file
-speechClient.recognize(fileName, options, (err, result) => {
-  if (err) {
-    console.error(err);
-    return;
-  }
-
-  console.log(`Transcription: ${result}`);
-});
+speechClient.recognize(fileName, options)
+  .then((results) => {
+    const transcription = results[0];
+    console.log(`Transcription: ${transcription}`);
+  });
 // [END speech_quickstart]
diff --git a/speech/recognize.js b/speech/recognize.js
index be5beb4129..45112d48b9 100644
--- a/speech/recognize.js
+++ b/speech/recognize.js
@@ -23,54 +23,66 @@
 
 'use strict';
 
-const fs = require('fs');
-const record = require('node-record-lpcm16');
-const speech = require('@google-cloud/speech')();
+const Speech = require('@google-cloud/speech');
 
 // [START speech_sync_recognize]
-function syncRecognize (filename, callback) {
-  // Detect speech in the audio file, e.g. "./resources/audio.raw"
-  speech.recognize(filename, {
+function syncRecognize (filename) {
+  // Instantiates a client
+  const speech = Speech();
+
+  const config = {
+    // Configure these settings based on the audio you're transcribing
     encoding: 'LINEAR16',
     sampleRate: 16000
-  }, (err, results) => {
-    if (err) {
-      callback(err);
-      return;
-    }
+  };
 
-    console.log('Results:', results);
-    callback();
-  });
+  // Detects speech in the audio file, e.g. "./resources/audio.raw"
+  return speech.recognize(filename, config)
+    .then((results) => {
+      const transcription = results[0];
+      console.log(`Transcription: ${transcription}`);
+      return transcription;
+    });
 }
 // [END speech_sync_recognize]
 
 // [START speech_async_recognize]
-function asyncRecognize (filename, callback) {
-  // Detect speech in the audio file, e.g. "./resources/audio.raw"
-  speech.startRecognition(filename, {
+function asyncRecognize (filename) {
+  // Instantiates a client
+  const speech = Speech();
+
+  const config = {
+    // Configure these settings based on the audio you're transcribing
     encoding: 'LINEAR16',
     sampleRate: 16000
-  }, (err, operation) => {
-    if (err) {
-      callback(err);
-      return;
-    }
+  };
 
-    operation
-      .on('error', callback)
-      .on('complete', (results) => {
-        console.log('Results:', results);
-        callback();
-      });
-  });
+  // Detects speech in the audio file, e.g. "./resources/audio.raw"
+  // This creates a recognition job that you can wait for now, or get its result
+  // later.
+  return speech.startRecognition(filename, config)
+    .then((results) => {
+      const operation = results[0];
+      // Get a Promise represention the final result of the job
+      return operation.promise();
+    })
+    .then((transcription) => {
+      console.log(`Transcription: ${transcription}`);
+      return transcription;
+    });
 }
 // [END speech_async_recognize]
 
 // [START speech_streaming_recognize]
+const fs = require('fs');
+
 function streamingRecognize (filename, callback) {
+  // Instantiates a client
+  const speech = Speech();
+
   const options = {
     config: {
+      // Configure these settings based on the audio you're transcribing
       encoding: 'LINEAR16',
       sampleRate: 16000
     }
@@ -90,9 +102,15 @@ function streamingRecognize (filename, callback) {
 // [END speech_streaming_recognize]
 
 // [START speech_streaming_mic_recognize]
-function streamingMicRecognize (filename) {
+const record = require('node-record-lpcm16');
+
+function streamingMicRecognize () {
+  // Instantiates a client
+  const speech = Speech();
+
   const options = {
     config: {
+      // Configure these settings based on the audio you're transcribing
       encoding: 'LINEAR16',
       sampleRate: 16000
     }
@@ -110,43 +128,39 @@ function streamingMicRecognize (filename) {
 }
 // [END speech_streaming_mic_recognize]
 
-// The command-line program
-var cli = require('yargs');
-var utils = require('../utils');
-
-var program = module.exports = {
-  syncRecognize: syncRecognize,
-  asyncRecognize: asyncRecognize,
-  streamingRecognize: streamingRecognize,
-  streamingMicRecognize: streamingMicRecognize,
-  main: function (args) {
-    // Run the command-line program
-    cli.help().strict().parse(args).argv;
-  }
-};
-
-cli
+require(`yargs`)
   .demand(1)
-  .command('sync <filename>', 'Detects speech in an audio file.', {}, function (options) {
-    program.syncRecognize(options.filename, utils.makeHandler(false));
-  })
-  .command('async <filename>', 'Creates a job to detect speech in an audio file, and waits for the job to complete.', {}, function (options) {
-    program.asyncRecognize(options.filename, utils.makeHandler(false));
-  })
-  .command('stream <filename>', 'Detects speech in an audio file by streaming it to the Speech API.', {}, function (options) {
-    program.streamingRecognize(options.filename, utils.makeHandler(false));
-  })
-  .command('listen', 'Detects speech in a microphone input stream.', {}, function () {
-    program.streamingMicRecognize();
-  })
-  .example('node $0 sync ./resources/audio.raw', 'Detects speech in "./resources/audio.raw".')
-  .example('node $0 async ./resources/audio.raw', 'Creates a job to detect speech in "./resources/audio.raw", and waits for the job to complete.')
-  .example('node $0 stream ./resources/audio.raw', 'Detects speech in "./resources/audio.raw" by streaming it to the Speech API.')
-  .example('node $0 listen', 'Detects speech in a microphone input stream.')
+  .command(
+    `sync <filename>`,
+    `Detects speech in an audio file.`,
+    {},
+    (opts) => syncRecognize(opts.filename)
+  )
+  .command(
+    `async <filename>`,
+    `Creates a job to detect speech in an audio file, and waits for the job to complete.`,
+    {},
+    (opts) => asyncRecognize(opts.filename)
+  )
+  .command(
+    `stream <filename>`,
+    `Detects speech in an audio file by streaming it to the Speech API.`,
+    {},
+    (opts) => streamingRecognize(opts.filename, () => {})
+  )
+  .command(
+    `listen`,
+    `Detects speech in a microphone input stream.`,
+    {},
+    streamingMicRecognize
+  )
+  .example(`node $0 sync ./resources/audio.raw`)
+  .example(`node $0 async ./resources/audio.raw`)
+  .example(`node $0 stream ./resources/audio.raw`)
+  .example(`node $0 listen`)
   .wrap(120)
   .recommendCommands()
-  .epilogue('For more information, see https://cloud.google.com/speech/docs');
-
-if (module === require.main) {
-  program.main(process.argv.slice(2));
-}
+  .epilogue(`For more information, see https://cloud.google.com/speech/docs`)
+  .help()
+  .strict()
+  .argv;
diff --git a/speech/system-test/quickstart.test.js b/speech/system-test/quickstart.test.js
index 1dbc4b5af1..1b24b541f4 100644
--- a/speech/system-test/quickstart.test.js
+++ b/speech/system-test/quickstart.test.js
@@ -26,33 +26,33 @@ const config = {
 };
 
 describe(`speech:quickstart`, () => {
-  let speechMock, SpeechMock;
-
   it(`should detect speech`, (done) => {
     const expectedFileName = `./resources/audio.raw`;
     const expectedText = `how old is the Brooklyn Bridge`;
 
-    speechMock = {
-      recognize: (_fileName, _config, _callback) => {
+    const speechMock = {
+      recognize: (_fileName, _config) => {
         assert.equal(_fileName, expectedFileName);
         assert.deepEqual(_config, config);
-        assert.equal(typeof _callback, `function`);
-
-        speech.recognize(fileName, config, (err, transcription, apiResponse) => {
-          _callback(err, transcription, apiResponse);
-          assert.ifError(err);
-          assert.equal(transcription, expectedText);
-          assert.notEqual(apiResponse, undefined);
-          assert.equal(console.log.calledOnce, true);
-          assert.deepEqual(console.log.firstCall.args, [`Transcription: ${expectedText}`]);
-          done();
-        });
+
+        return speech.recognize(fileName, config)
+          .then((results) => {
+            const transcription = results[0];
+            assert.equal(transcription, expectedText);
+
+            setTimeout(() => {
+              assert.equal(console.log.callCount, 1);
+              assert.deepEqual(console.log.getCall(0).args, [`Transcription: ${expectedText}`]);
+              done();
+            }, 200);
+
+            return results;
+          });
       }
     };
-    SpeechMock = sinon.stub().returns(speechMock);
 
     proxyquire(`../quickstart`, {
-      '@google-cloud/speech': SpeechMock
+      '@google-cloud/speech': sinon.stub().returns(speechMock)
     });
   });
 });
diff --git a/speech/system-test/recognize.test.js b/speech/system-test/recognize.test.js
index 8249321d2c..2653725fd3 100644
--- a/speech/system-test/recognize.test.js
+++ b/speech/system-test/recognize.test.js
@@ -25,14 +25,14 @@ const text = `how old is the Brooklyn Bridge`;
 
 describe(`speech:recognize`, () => {
   it(`should run sync recognize`, () => {
-    assert.equal(run(`${cmd} sync ${filename}`, cwd), `Results: ${text}`);
+    assert.equal(run(`${cmd} sync ${filename}`, cwd), `Transcription: ${text}`);
   });
 
   it(`should run async recognize`, () => {
-    assert.equal(run(`${cmd} async ${filename}`, cwd), `Results: ${text}`);
+    assert.equal(run(`${cmd} async ${filename}`, cwd), `Transcription: ${text}`);
   });
 
   it(`should run streaming recognize`, () => {
-    assert.notEqual(run(`${cmd} stream ${filename}`, cwd).indexOf(text), -1);
+    assert.equal(run(`${cmd} stream ${filename}`, cwd).includes(text), true);
   });
 });
diff --git a/speech/test/quickstart.test.js b/speech/test/quickstart.test.js
deleted file mode 100644
index 807e23e507..0000000000
--- a/speech/test/quickstart.test.js
+++ /dev/null
@@ -1,49 +0,0 @@
-/**
- * Copyright 2016, Google, Inc.
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *    http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-'use strict';
-
-const proxyquire = require(`proxyquire`).noCallThru();
-
-const config = {
-  encoding: 'LINEAR16',
-  sampleRate: 16000
-};
-
-describe(`speech:quickstart`, () => {
-  let speechMock, SpeechMock;
-  const error = new Error(`error`);
-  const fileName = `./resources/audio.raw`;
-
-  before(() => {
-    speechMock = {
-      recognize: sinon.stub().yields(error)
-    };
-    SpeechMock = sinon.stub().returns(speechMock);
-  });
-
-  it(`should handle error`, () => {
-    proxyquire(`../quickstart`, {
-      '@google-cloud/speech': SpeechMock
-    });
-
-    assert.equal(SpeechMock.calledOnce, true);
-    assert.deepEqual(SpeechMock.firstCall.args, [{ projectId: 'YOUR_PROJECT_ID' }]);
-    assert.equal(speechMock.recognize.calledOnce, true);
-    assert.deepEqual(speechMock.recognize.firstCall.args.slice(0, -1), [fileName, config]);
-    assert.equal(console.error.calledOnce, true);
-    assert.deepEqual(console.error.firstCall.args, [error]);
-  });
-});
diff --git a/speech/test/recognize.test.js b/speech/test/recognize.test.js
deleted file mode 100644
index 89747c0245..0000000000
--- a/speech/test/recognize.test.js
+++ /dev/null
@@ -1,40 +0,0 @@
-/**
- * Copyright 2016, Google, Inc.
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *    http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */
-
-'use strict';
-
-const proxyquire = require(`proxyquire`).noCallThru();
-
-describe(`speech:recognize`, () => {
-  it(`should handle errors`, () => {
-    const filename = `audio.raw`;
-    const error = new Error(`error`);
-    const callback = sinon.spy();
-    const speechMock = {
-      recognize: sinon.stub().yields(error),
-      startRecognition: sinon.stub().yields(error)
-    };
-    const SpeechMock = sinon.stub().returns(speechMock);
-    const program = proxyquire(`../recognize`, {
-      '@google-cloud/speech': SpeechMock
-    });
-
-    program.syncRecognize(filename, callback);
-    program.asyncRecognize(filename, callback);
-
-    assert.equal(callback.callCount, 2);
-    assert.equal(callback.alwaysCalledWithExactly(error), true);
-  });
-});