Setup guide
Set up Azure Speech
Microsoft's Azure Speech service has high-quality neural voices in a huge range of languages, and its free tier renews every month, which makes it a good fit for regular listening. This guide walks you through it step by step. It takes about 10 minutes.
What you'll need
- A Microsoft account and a free Azure account (a payment method is required to sign up)
- About 10 minutes
-
1
Create a free Azure account
If you don't have an Azure account yet:
- Go to azure.microsoft.com/free
- Click Start free and sign in with (or create) a Microsoft account
- Complete identity verification and add a payment method
Nothing is charged on the free (F0) Speech tier. The card is only used for identity verification and for paid tiers you explicitly choose.
-
2
Create a Speech resource
- Sign in to the Azure Portal
- Click Create a resource and search for Speech
- Select Speech (by Microsoft, under "AI + Machine Learning") and click Create
- Choose your subscription and create (or pick) a resource group
- Pick a Region close to you, e.g. eastus, westeurope, or southeastasia
- Give the resource a name, e.g. cloud-speech-tts
- For Pricing tier, choose Free (F0); you can upgrade to Standard (S0) later if you need more
- Click Review + create, then Create
Remember the region you picked. The extension needs it in the next step.
-
3
Copy your key and region
- Once deployment finishes, click Go to resource
- In the left menu, open Keys and Endpoint (under "Resource Management")
- Copy Key 1 (either key works)
- Note the Location/Region shown on the same page, e.g. eastus
Security warning
Treat the key like a password; anyone with it can use your Speech resource. You can regenerate it from the same page at any time. -
4
Connect the extension
- Click the extension icon in your browser toolbar
- Open Settings → Azure Speech
- Paste your Key
- Enter your Region (e.g. eastus)
- Click Save & test
Once the test passes, the Azure voices appear in the voice picker. Press ▶ to preview any voice and ★ to favorite the ones you like. Then highlight text on any page, right-click, and choose Read aloud, or press Ctrl/Cmd+Shift+S.
Costs
The Free (F0) tier includes about 0.5 million neural characters per month, and it renews every month, not just for the first year. On the Standard (S0) tier, neural voices cost roughly $15-16 per 1 million characters; HD and custom voices cost more. Prices vary by region and change over time, so check the official Azure Speech pricing page and our provider comparison.
Remove access when you're done
- In the Azure Portal, search for your Speech resource (or open its resource group).
- Select the resource and choose Delete. Both keys stop working immediately. Deleting the whole resource group removes everything it contains.
- In the extension, open Settings → Azure Speech and clear the saved key (or turn the provider off).
Prefer the command line?
With the Azure CLI, this creates the same free-tier Speech resource as the steps above:
az login
az group create --name tts --location eastus
az cognitiveservices account create --name my-speech --resource-group tts \
--kind SpeechServices --sku F0 --location eastus --yes
az cognitiveservices account keys list --name my-speech --resource-group tts
# Paste key1 into the extension, with region "eastus".
And to remove it all later:
az cognitiveservices account delete --name my-speech --resource-group tts
az group delete --name tts --yes
Troubleshooting
"Credentials are missing or invalid"
Double-check the values you pasted; a stray space or missing character is the most common cause. Confirm the region matches your Speech resource (e.g. eastus) and that the key wasn't regenerated. Then click Save & test again.
"Access denied" errors
Your key is valid but lacks permission or quota. Verify the resource wasn't moved or deleted, your account is in good standing, and you haven't exceeded the F0 tier's monthly quota.
For no audio, missing voices, or keyboard shortcut problems, see the troubleshooting guide.
Prefer a different provider? See Amazon Polly, Google Cloud TTS, or OpenAI.