Async Python client for controlling Glen Dimplex Heating & Ventilation (GDHV) appliances via the Dimplex cloud API.
Important
Unofficial library. Not affiliated with, endorsed by, or supported by Dimplex, Glen Dimplex Heating & Ventilation (GDHV), or the Glen Dimplex Group. This is an independent project built by reverse-engineering the official Dimplex Control Android app and the private cloud API it talks to.
There is no public or documented API. Dimplex can change or withdraw that API at any time and without notice, which may break this library. Use it at your own risk.
No Dimplex code is included or redistributed. "Dimplex", "Quantum", "QRAD" and related marks belong to their respective owners and are used here only to describe compatibility.
dimplex-controller-py is an asynchronous Python client that talks to the GDHV IoT cloud platform. It handles Azure B2C authentication (including automatic token refresh), discovers your Hubs, Zones and Appliances, and lets you read telemetry and send control commands — all from a script or a larger application.
It is the engine behind the Dimplex Hub Home Assistant integration and is published to PyPI as dimplex-controller.
- Features
- Installation
- Quick start
- Usage guide
- Configuration
- API reference
- Troubleshooting
- Contributing
- Changelog
- Authentication — Azure B2C login with automatic token refresh and secure token persistence.
- Discovery — List Hubs, Zones and Appliances linked to your account.
- Real-time status — Fetch room temperature, setpoints, comfort status, boost/away modes and EcoStart state.
- Control — Set operation modes, activate Boost and Away, toggle EcoStart and Open Window Detection, and programme timer schedules.
- Energy telemetry — Pull Time Series Insights (TSI) energy reports with a robust telemetry parser that adapts to varying firmware formats.
pip install dimplex-controllergit clone https://github.com/KRoperUK/dimplex-controller-py.git
cd dimplex-controller-py
pip install .git clone https://github.com/KRoperUK/dimplex-controller-py.git
cd dimplex-controller-py
uv syncuv sync creates .venv, installs the library in editable mode, and installs the
dev dependency group from uv.lock.
Requires: Python 3.10 or later.
The library uses asyncio and aiohttp. Here is the smallest example that lists your Hubs and Zones:
import asyncio
from aiohttp import ClientSession
from dimplex_controller import DimplexControl
async def main() -> None:
async with ClientSession() as session:
client = DimplexControl(session, refresh_token="YOUR_REFRESH_TOKEN")
hubs = await client.get_hubs()
for hub in hubs:
print(f"Hub: {hub.Name}")
zones = await client.get_hub_zones(hub.HubId)
for zone in zones:
print(f" Zone: {zone.ZoneName}")
if __name__ == "__main__":
asyncio.run(main())Dimplex uses Azure AD B2C. The library supports two methods:
client = DimplexControl(session)
await client.auth.headless_login("you@example.com", "password")This automates the full B2C flow via HTTP. On success, client.is_authenticated is True and tokens can be persisted with client.export_tokens().
Run demo.py to open a browser, sign in, and paste the redirect URL. The script saves tokens to dimplex_tokens.json. Subsequent runs load the refresh token automatically.
Either way, refresh tokens are used on future calls — the library handles token renewal transparently.
hubs = await client.get_hubs()
for hub in hubs:
print(f"Hub: {hub.Name} ({hub.HubId})")
zones = await client.get_hub_zones(hub.HubId)
for zone in zones:
print(f" Zone: {zone.ZoneName} ({zone.ZoneId})")
appliances = zone.Appliances
for appliance in appliances:
print(f" Appliance: {appliance.ApplianceId}")from dimplex_controller.models import ApplianceStatus
status_list = await client.get_appliance_overview(hub_id, [appliance_id_1, appliance_id_2])
for status in status_list:
print(f"Room temperature: {status.RoomTemperature}°C")
print(f"Target temperature: {status.ActiveSetPointTemperature}°C")
print(f"EcoStart enabled: {status.EcoStartEnabled}")
print(f"Comfort status: {status.ComfortStatus}")A note on empty responses: when every requested appliance is offline (e.g. radiators switched off at the wall) the cloud returns HTTP 200 with an empty list.
get_appliance_overviewsurfaces that as[]— it is not an error. If you need a stable id → status mapping, useget_appliance_overview_map(...), which fills inNonefor missing ids.
# Enable EcoStart
await client.set_eco_start(hub_id, [appliance_id], True)
# Enable Open Window Detection
await client.set_open_window_detection(hub_id, [appliance_id], True)
# Timed Boost (ApplianceModes=2, Time = minutes)
await client.set_boost(hub_id, [appliance_id], temperature=25.0, duration_minutes=60)
# Away until a given moment (ApplianceModes=4). Away is a settable 7–18 °C
# setback and defaults to the 7 °C anti-freeze floor. A higher temperature is
# clamped to 18 °C with a warning, matching what the cloud does to it anyway.
from datetime import datetime, timedelta, timezone
await client.set_away(
hub_id,
[appliance_id],
temperature=12.0,
until=datetime.now(timezone.utc) + timedelta(days=3),
)
# Set the target temperature — dedicated endpoint, leaves the schedule alone
await client.set_appliance_setpoint_temperature(hub_id, [appliance_id], 21.5)
# Turn off the way the app does: frost protection at 7 °C
await client.turn_off(hub_id, [appliance_id])Mode flag values matter.
EApplianceModesis a bitfield where Boost is2and Away is4;16is Advance and32is FrostProtect. Releases before 0.13.0 had these wrong, so Boost silently became Advance and Away became a fixed 7 °C frost hold. If you hand-buildApplianceModeSettings, useApplianceModeFlag. See docs/decompiled-api-reference.md.Avoid
SetTimerModefor control.set_mode()and the deprecatedset_target_temperature()rewrite the schedule; Quantum rejects that with HTTP 403. Useset_appliance_setpoint_temperature()andturn_off().
from dimplex_controller import parse_telemetry_points, summarise_energy
report = await client.get_tsi_energy_report(hub_id)
for appliance_id, telemetry in report.ApplianceTelemetryData.items():
points = parse_telemetry_points(telemetry)
daily = summarise_energy(points, mode="daily")
lifetime = summarise_energy(points, mode="lifetime")
print(f"{appliance_id}: today={daily.total_kwh} kWh, lifetime={lifetime.total_kwh} kWh")parse_telemetry_points normalises firmware-varying point shapes. summarise_energy builds daily (local midnight) and lifetime totals per register. T1 (off-peak / cheaper) and T2 (peak / more expensive) must not be summed; parse with VALUE_KEY_T1 / VALUE_KEY_T2. With include_previous_period=True the cloud often returns full history — filter client-side rather than trusting days_back alone.
See docs/compatibility.md for the library ↔ Home Assistant version matrix.
| Environment variable | Purpose |
|---|---|
DIMPLEX_TOKENS_FILE |
Path to the JSON token store. Defaults to dimplex_tokens.json. |
Main client class. Construct with an aiohttp.ClientSession and a refresh_token (or token_bundle).
| Method | Description |
|---|---|
get_hubs() |
Returns list[Hub]. |
get_hub_zones(hub_id) |
Returns list[Zone] for a Hub. |
get_zone(hub_id, zone_id) |
Returns a single Zone. |
get_appliance_overview(hub_id, appliance_ids) |
Returns list[ApplianceStatus] (may be []). |
get_appliance_overview_map(hub_id, appliance_ids) |
Stable dict[str, ApplianceStatus | None]. |
get_user_context() |
Returns UserContext. |
get_product_models() |
Returns list[ProductModel] (cacheable). |
get_schedule(hub_id, appliance_id) |
Returns TimerModeSettings (timer + periods). |
set_mode(hub_id, appliance_id, mode) |
Rewrite TimerMode. 403 on Quantum — prefer turn_off. |
set_target_temperature(hub_id, appliance_id, temp) |
Deprecated: rewrites period setpoints. Prefer set_appliance_setpoint_temperature. |
set_appliance_setpoint_temperature(hub_id, appliance_ids, temperature) |
Preferred setpoint path; non-destructive. |
set_period_setpoint(...) |
Update one timer period without clobbering siblings. |
update_period(...) |
Replace a timer period matched by day + start time. |
copy_schedule_to_appliances(...) |
Copy one appliance's schedule onto others. |
set_boost(hub_id, appliance_ids, *, temperature, duration_minutes, enable) |
Timed Boost (ApplianceModes=2). |
clear_boost(hub_id, appliance_ids) |
Disable Boost. |
set_away(hub_id, appliance_ids, *, temperature, enable, until, number_of_days) |
Away setback (ApplianceModes=4). |
clear_away(hub_id, appliance_ids) |
Disable Away. |
set_frost_protect(...) / turn_off(...) |
Frost protection at 7 °C — the app's "off". |
set_advance(...) |
Advance to the next schedule period. |
set_manual(...) / set_eco_mode(...) |
Manual / Eco mode holds. |
set_setback_temperature(...) |
Write the setback temperature (untested). |
set_eco_start(hub_id, appliance_ids, enable) |
Toggle EcoStart. |
set_open_window_detection(hub_id, appliance_ids, enable) |
Toggle Open Window Detection. |
set_hot_water_*(...) / *_heat_pump_hot_water_schedule(...) |
Hot-water cylinder surface (untested — no hardware). |
get_tsi_energy_report(hub_id, ...) |
Returns TsiEnergyReport. |
capabilities_for(appliance, *, status, product) |
Derive an ApplianceCapabilities matrix. |
export_tokens() / apply_tokens(bundle) |
Token persistence helpers. |
Hub— Hub metadata.Zone— Zone metadata with linked Appliances.Appliance— Appliance metadata.ApplianceStatus— Live telemetry (room temperature, setpoints, comfort, etc.).ApplianceModeSettings— Payload for mode changes.TimerPeriod/TimerModeSettings— Timer schedule structures.UserContext— Authenticated user profile.TsiEnergyReport— Energy telemetry keyed by appliance.
DimplexError— Base exception.DimplexAuthError— Authentication or token errors.DimplexApiError— API returned a non-success status. Containsstatusandmessage.DimplexConnectionError— Network-level failures.
- Verify that the refresh token in
dimplex_tokens.jsonhas not expired. Delete the file and re-rundemo.pyto capture a fresh one. - Ensure your network can reach
login.microsoftonline.comand the Dimplex API endpoints. - If you have multi-factor authentication (MFA) enabled on your Dimplex account, the headless flow should still work because it uses a browser session you control manually.
This means the API rejected the token. Common causes:
- Token file is missing or corrupt.
- The refresh token has expired (Azure B2C refresh tokens typically last 90 days).
- The token was revoked from the Azure portal.
Fix: Delete dimplex_tokens.json and re-run the demo.py flow.
The library could not reach the GDHV API. Check:
- Internet connectivity.
- DNS resolution for
api.gdhv.io(or whatever endpoint is configured inconst.py). - No corporate firewall or proxy is blocking
HTTPStraffic.
If parse_telemetry_points returns an empty list, the API likely returned an unexpected schema for your firmware version. Please open an issue with a redacted example of the raw response so the parser can be updated.
The GDHV cloud API has rate limits. The library retries idempotent GET
requests automatically on HTTP 429/5xx and connection errors, using exponential
backoff with jitter and honouring the Retry-After header when present.
Non-idempotent control calls (POST/PUT/PATCH/DELETE) are not retried
by default. Tune this via the client constructor:
client = DimplexControl(
session,
refresh_token="...",
max_retries=3, # retries after the first attempt (0 disables)
retry_base_delay=0.5, # seconds; exponential base
retry_max_delay=8.0, # seconds; backoff ceiling
retry_non_idempotent=False, # set True to also retry POST/PUT/etc.
)If you still hit persistent limits, back off for a few minutes before retrying.
This is the cloud's normal response when every requested appliance is offline (e.g. radiators turned off at the wall, or a hub that has dropped off the network). It is not an error — get_appliance_overview returns [] and get_appliance_overview_map returns a dict of None values. Treat the call as a successful poll; the appliances will reappear in subsequent calls once they come back online. See the note in Reading status for details.
Install the package (or an editable install) to get the dimplex console script:
pip install dimplex-controller
export DIMPLEX_REFRESH_TOKEN=... # never commit this
dimplex login
dimplex hubs
dimplex zones --hub <hub-id> -v
dimplex status <hub-id> <appliance-id>
dimplex energy <hub-id> --days 30
# control writes require --yes
dimplex boost <hub-id> <appliance-id> --minutes 60 --yesTokens can also come from a JSON file (--tokens-file / DIMPLEX_TOKENS_FILE) with keys refresh_token, access_token, expires_at. Secrets are redacted in CLI output unless --show-tokens is passed.
Pull requests into main must keep the ci GitHub Actions check green.
- Changes under
dimplex_controller/,tests/, or CI config run lint, pre-commit, and the pytest matrix (Python 3.10–3.13). Thecijob fails if any of those fail. - Docs-only PRs still report a green
ciwithout running the full matrix.
Direct pushes to main are blocked (PR + squash only; no force-push/delete). Commits must be signed (repo-wide rule).
Contributions are welcome! Please read the contributing guidelines before opening a pull request.
Key points:
- Use Conventional Commits (
feat:,fix:,chore:, etc.) — this drives the automated changelog and PyPI releases. - Run
ruff check,ruff format --checkandpytestlocally before pushing. - Pre-commit hooks are available — run
pre-commit installonce.
See CHANGELOG.md for version history.