Amazon Polly smart home integration, free and open source
Amazon Polly voices for Gladys announcements, with your own AWS account.
Published by LM1LC3N7
- Devices
- Version 0.1.1
- Requires Gladys >=5.3.0
- ★ 0
Gives Gladys the voices of Amazon Polly for its announcements (the "Play a notification" scene action) and the voice assistant reply, with your own AWS account. Unofficial: not affiliated with Amazon.
This integration has never been tested with a real AWS account yet: it was built from the Amazon Polly documentation and tested against a local imitation of the Polly API.
What it does
Gladys sends a text (2,000 characters at most) and the language of the user. The integration picks the voice you chose for that language (French, English or German; the "Language when unknown" voice otherwise), asks Amazon Polly for an MP3 clip and hands it to Gladys, which plays it on your speakers.
Then, in Gladys, open Integrations → Artificial Intelligence, card Voice, and select Amazon Polly.
Limits of Gladys itself (not of this integration): the clip is served to speakers on your local network, and the voice assistant reply stays in text in an HTTPS page (see the Gladys documentation).
What it costs
Amazon Polly is billed per character sent, on your AWS account (prices on aws.amazon.com/polly/pricing, November 2026: about 4 USD per million characters with the standard engine, about 16 USD with neural, more with generative and long-form; new accounts get a free allowance during their first year). The standard engine is the default.
To keep the bill under control:
- Monthly limit (default 1,000,000 characters): counted in memory before each request leaves, and written to
/data(the write waited for 1 s at most); once reached, nothing more is sent until the next calendar month (UTC) or a higher limit. A request that may have been billed (no answer, connection cut) stays counted; a request AWS answered with an error is not. An announcement is never blocked by/data(delayed 3 s at most): when it is not read yet (2 s at most), read-only or slow, the request leaves all the same and the status says "the monthly counter is not saved: the limit is only guaranteed until the next restart" (the pause and a refusal kept there may be unknown too). A cut of power or a crash at the wrong moment can lose the count of one request. - The same text in the same voice is played again from memory (24 hours, 20 clips), not billed twice; asked several times at once (a scene announcing on several speakers), it makes one request whose answer all share.
- At most 2 requests at once and 8 waiting; more at the same moment are refused ("Too many announcements").
- No retry: a failed request is not sent again by the integration (a retry could be billed twice).
Create the AWS key (minimal rights)
- In the AWS console, IAM → Users → Create user (no console access).
- Attach this policy only (it allows the synthesis, nothing else):
{
"Version": "2012-10-17",
"Statement": [{ "Effect": "Allow", "Action": "polly:SynthesizeSpeech", "Resource": "*" }]
}
- Security credentials → Create access key ("Application running outside AWS"). Copy the access key ID
(
AKIA…) and the secret access key (shown once). - Optionally, set an AWS Budget alert on your account: it is the only limit AWS enforces on its side.
Configuration
| Field | What it does |
|---|---|
| Access key ID | The AKIA… identifier of the IAM user. |
| Secret access key | Kept by Gladys in its database, never shown again, sent only to Amazon Web Services: it signs each request; only the signature travels, not the key itself. |
| AWS region | Where Polly runs (default eu-west-1, Ireland). Not every engine exists in every region. |
| Voice engine | standard (default, cheapest), neural, generative, long-form. |
| French / English / German voice | The voice for each language. Voices marked "not standard" need the neural engine or a newer one: Polly refuses them with the standard engine, and the error says so. |
| Language when unknown | The voice used when Gladys sends no language or another one. |
| Speaking rate | Very slow to very fast (SSML prosody; the tags are not billed). |
| Monthly limit (characters) | 1,000 to 100,000,000; default 1,000,000. |
The Test the voice button asks Polly for a short sentence (about 50 billed characters) with the saved settings and says how long it took. Nothing is played.
When something goes wrong
The status of the Configuration screen says why:
- AWS refused these keys (unknown, deleted or deactivated key, wrong secret, or a region not enabled on
the account, such as eu-south-2 or af-south-1): nothing more is sent with them in that region, even after a restart (once
/datais read), until you enter other keys, choose another region, or the Test button succeeds (it sends anyway)./datakeeps only 8 bits of a salted hash of the refused keys, never the keys: after a restart, other keys have 1 chance in 256 to look refused; the Test button then clears it. - Denied (
AccessDeniedException): the keys are valid but the IAM user lackspolly:SynthesizeSpeech. This is not a refusal of the keys: the next announcement tries again. - Clock: AWS rejected the time of the request; check the clock of the Gladys host.
- Slow down / Polly failed (throttling, HTTP 429 or 5xx): no request for 60 s, then 2, 4… minutes up to
1 hour if it goes on (longer when AWS asks for it with
Retry-After, 6 hours at most); the Test button waits too. The pause is kept in/dataacross a restart; a success ends it. - Polly refused the request: its reason is shown (for example "This voice does not support the selected engine"), cut before the first value it quotes (your text could be one, even in part); nothing is billed.
- Could not be reached, or the connection was cut: a cut after the request left stays counted.
- Did not answer in time: Gladys waits 30 s; the integration gives up at 25 s.
Privacy
The text to speak is sent to Amazon Web Services in the region you chose. The integration never writes it to
the logs or to /data (only its length); the clips stay in memory only.
Configuration settings
These are the settings Amazon Polly asks for in its configuration screen in Gladys.
| Setting | Type | Required | Description |
|---|---|---|---|
| Your AWS keys | section | No | Amazon Polly is billed to your own AWS account, per character spoken. Create an IAM user allowed only polly:SynthesizeSpeech (the documentation gives the policy) and enter its access key here. The standard engine is the default: the cheapest one. |
| Access key ID | string | Yes | The access key ID of the IAM user (AKIA… ). |
| Secret access key | secret | Yes | The secret access key shown once by AWS when the key was created. Kept by Gladys in its database, never shown again, sent only to Amazon Web Services (it signs the requests: only the signature travels, not the key itself). |
| AWS region | select | No | Where Polly runs. Not every engine exists in every region: eu-west-1 (Ireland) has the standard and neural voices. |
| Voice engine | select | No | Standard is the cheapest; neural, generative and long-form sound more natural and cost more per character (see the prices). Not every voice has every engine: Polly then says so. |
| French voice | select | No | Used when the request is in this language. "not standard": this voice needs the neural or a newer engine. |
| English voice | select | No | Used when the request is in this language. "not standard": this voice needs the neural or a newer engine. |
| German voice | select | No | Used when the request is in this language. "not standard": this voice needs the neural or a newer engine. |
| Language when unknown | select | No | The voice used when Gladys sends no language, or another one. |
| Speaking rate | select | No | |
| Monthly limit (characters) | number | No | Nothing more is sent to Polly once this many characters were sent in the calendar month (UTC). 1,000,000 characters cost about 4 USD with the standard engine. |
How to install Amazon Polly in Gladys
- In Gladys, open Integrations: Amazon Polly appears in the catalog, next to the native integrations, with a community badge.
- Click Install. Gladys pulls the Docker image (
ghcr.io/lm1lc3n7/gladys-amazon-polly:0.1.1), starts it in a sandbox isolated from the core, and generates the integration's interface (devices, discovery and configuration). - Open the Configuration screen of the integration, fill in the settings, and save.
- You can also install it directly from its repository URL: https://github.com/LM1LC3N7/gladys-amazon-polly.
Amazon Polly requires Gladys >=5.3.0. The catalog inside Gladys refreshes every hour, so a new version becomes available at most one hour after its release.
Not running Gladys yet? It is free and open source: follow the installation guide to get started.
About external integrations
Amazon Polly is an external integration: a community integration packaged as a Docker container and published on GitHub, that Gladys installs in one click and runs in a sandbox isolated from its core. It is published and maintained by LM1LC3N7, not by the Gladys core team.
- Browse all integrations, native and community
- Build and publish your own external integration
- Source code on GitHub — source of this documentation