DIY vs proprietary grow platforms
What the commercial platforms actually cost, set against a DIY build with no subscription — and an honest account of what you give up by not paying them. Every figure carries the confidence level and source it came from.
Vendor research dated 2026-08-02. All vendor figures USD. DIY figures derived from this repository's own build recipes, converted at 1 NZD = 0.5879 USD (rate date 2026-08-01).
Read the tags before the numbers
(a) vendor-published · (b) reseller listing · (c) user-reported · (d) not publicly available. A vendor-published figure can still be MEDIUM confidence if it is a modelling default rather than a price list — which is exactly the case for AROYA's hardware anchors. Modelled numbers are not presented with the same weight as published ones.
Break-even, not a single number
A subscription is a recurring cost and a DIY build is a one-off, so a single total is misleading. The question that matters is how long until the platform costs more than building it yourself — and for some vendors and scales, the answer is never.
Where this comparison is and is not fair
Vendor build totals are arithmetic on tagged list prices, not vendor quotes. They exclude tax, shipping, electrical work and, except where stated, installation. These systems are also not functionally identical: a $925 environment controller and a $47,200 substrate-analytics platform with human agronomy coaching are not the same purchase. The tables compare cost. The honest-limits section compares capability.
The DIY side, derived from the catalogue
The DIY figures are not entered by hand. They are computed from the build recipes in this repository, which are themselves priced from the parts catalogue — so if a part price changes, this page changes with it. $838 per room is the PoE room controller plus a substrate sensing array.
Recurring cost: none. There is no subscription, no API fee, no per-seat charge and no per-sensor charge. That is the entire structural argument for DIY, and the rest of this page is about what it costs you elsewhere.
The vendors
TrolMaster Runs the room
Environment control. Dominant at small and mid indoor scale.
Environment control with relay/device outputs. Runs the room.
- Subscription
- App is free with no paid remote tier. API access is billed separately at $15 per month per device.
- Recurring cost
- "No subscription" is true for using it. The moment you want programmatic access to your own data it is $15 per device per month, indefinitely. At 20 devices that is $3,600/yr - the same order as AROYA's mid tiers.
- Stop paying — what happens?
- ANSWERED. Local control at the controller face is free and always available, with local CSV logging to MicroSD. Confirmed surviving a cloud outage. The catch is that alerts die with the cloud. (a)
- 4-room figure
- $6,391 hardware , no recurring (b) HIGH
Pulse Grow Monitoring only
Sensor-agnostic monitoring. The natural buy comparison for a DIY-inclined operator.
Monitoring and alerting only. No relay outputs, no valve or equipment control - Pulse cannot run a room on its own and is not a substitute for a controller.
- Subscription
- Working free tier. Enthusiast $10/user/month, Professional $35/user/month.
- Recurring cost
- Professional is effectively mandatory at four-room scale, because Enthusiast lacks multiple grow locations and Free lacks multi-user.
- Stop paying — what happens?
- ANSWERED. Vendor confirms hardware keeps working without payment. (a)
- 4-room figure
- $7,851 hardware , $840/yr recurring (a) HIGH
Growlink Runs the room
Direct AROYA competitor. Published SaaS tiers.
Irrigation and environment control with valve and device outputs.
- Subscription
- Root free, Sprout $25/month, Bloom $250/month per facility, Harvest $1,000/month per facility.
- Recurring cost
- Entry tier confirmed at $240-300/year from two independent vendor sources: a historical Pro Starter price of $20/month, and archived promo terms describing a 12-month Sprout subscription as "a $300 value".
- Stop paying — what happens?
- NOT ANSWERED. A full-text search of the current and archived Terms of Service found no "Effect of Termination" clause, no statement on whether the controller keeps running, no data export obligation and no wind-down period. The Terms do reserve the right to suspend access on non-payment. (d)
- 4-room figure
- $11,964 hardware
, $3,000/yr recurring
(a) MEDIUM
Published-list floor, not an obtainable quote. Growlink state that commercial operations over 32 sensors require a separate enterprise agreement.
AROYA (Addium) Analytics / agronomy
Category-defining in cannabis crop steering. Highest subscription exposure.
Substrate crop-steering analytics and agronomy coaching. Actuation generally comes from separate irrigation hardware.
- Subscription
- $150 to $3,900 per month, billed per canopy square foot, facility-wide. Room count is not the billing unit.
- Recurring cost
- Because billing is by canopy area rather than rooms, a "4-room AROYA price" does not exist as a concept. Two canopy scenarios are shown instead.
- Stop paying — what happens?
- DECLINED. AROYA's own FAQ answers the question with "There are a few factors... Please reach out to us to book a demo." (d)
- 4-room figure
- $35,000 hardware
+ $5,000 implementation
, $7,200/yr recurring
(a) MEDIUM
Split confidence. The subscription figure is HIGH - it is labelled by AROYA as real 2025 pricing. The hardware and implementation figures are MEDIUM: they are calculator modelling defaults (TIER_ANCHORS), not a confirmed list price.at approximately 8,000 sq ft canopy
Argus Controls Runs the room
Enterprise horticultural control. The only vendor with a published price list carrying real dollar figures.
Full engineered control system, configured by the vendor.
- Subscription
- Capex model. Titan software from $23,000 (Basic) or $28,000 (Enterprise, adding multi-user, audit tracking and password control).
- Recurring cost
- Ongoing costs are not published. Payment terms are 20% deposit, 80% pre-ship, roughly 12 weeks deposit to delivery, and the customer supplies the installation crew.
- Stop paying — what happens?
- Not applicable in the same way - this is a capex purchase rather than a subscription platform. (a)
- 4-room figure
- $51,000 hardware
, no recurring
(a) HIGH-STALE
Vendor-published, but from the 2018 Estimating Guide. These are 2018 dollars and should be treated as a floor, not a current quote.
Grow Sensor Monitoring only
Canopy-level optical and climate sensing. Leaf VPD, PPFD, spectrum and DLI.
Monitoring only. No relay outputs and no equipment control - it measures the canopy, it does not run the room.
- Subscription
- No subscription for the app. The vendor FAQ states "The Grow App is free to use with your Grow Sensor. Premium features, such as AI insights, may require a subscription in the future." The PRO bundle includes one year of the GrowOps crop-tracking platform; the price after that year is not published.
- Recurring cost
- Currently no recurring cost to use the sensor. Two forward-looking unknowns are stated by the vendor rather than inferred: premium features "may require a subscription in the future", and GrowOps renewal pricing after the bundled first year is not published.
- Stop paying — what happens?
- Not applicable at present - there is no subscription required to use the sensor or the app. (a)
- 4-room figure
- $3,770 hardware
, no recurring
(a) HIGH
Vendor Shopify product feed. Four PRO systems at GBP 699. Converted to USD using the cross-rate in data/fx.yaml.
Where pricing simply is not obtainable
Two vendors could not be priced at all. That is recorded here as a finding rather than filled in with an estimate — "not published anywhere, including through distributors" is itself decision-relevant.
Priva (d)
No pricing published anywhere, including on authorised reseller storefronts, which list products at zero with a "View Product Details" prompt. US procurement and grant-budget searches turned up no disclosed contract values. API access additionally requires a Priva Gateway and a Priva Connected subscription, whose price is also unpublished.
Relevance: Enterprise glass horticulture. A small indoor operator would not reach the quote stage economically.
iUNU (d)
No pricing published; everything routes to a demo booking or a gated form. Sold as system-as-a-service on 3-5 year SLAs with vendor-financed hardware, priced by camera passes per day.
Relevance: Has exited cannabis as a go-to-market. The "Crops Grown" field on iUNU's own contact form offers tomatoes, cucumbers, peppers, berries, leafy greens, ornamentals and other - cannabis is not an option (fetched August 2026). The product is also camera-on-rail for tall vine crops in glass, which is a poor structural fit for racked indoor rooms.
Costs no vendor publishes
These are real and material, and every one of them is (d). None appear in the tables above, on either side.
| Cost | What is known |
|---|---|
| Sensor replacement | TrolMaster warrants products for 3 years but sensor probes for 1 year, against a consistent body of user reports of drift and failure around the two-year mark. CO2 sensors run $187-243 each and substrate probes $200-309. No vendor publishes a failure rate. |
| Renewal increases | AROYA's contract grants repricing at the "then going rate" with 30 days' notice and no cap. Growlink auto-renews at the "then-current standard rate", also uncapped, and has moved price three times in 30 months. No user has publicly reported an actual renewal invoice. |
| Installation and commissioning | AROYA's calculator carries $3,000-$20,000 implementation. Growlink does not publish it, though one user reports approximately $15,000 install on a $60,000 hardware order. Argus require the customer to supply the installation crew. |
| Calibration labour | TEROS 12 is factory-calibrated and cannot be field-recalibrated - drift means RMA or replacement. TrolMaster substrate sensors need user-defined calibration to your specific media, reported drifting by 10-20% every few days. Pulse CO2 needs manual outdoor recalibration every 4-6 months and pH probes are a $99 consumable. None of this is a line item anywhere. |
| The audit sensor | Multiple experienced operators run a second trusted sensor purely to check the first. That is a real recurring cost of a system you do not fully trust, and it appears in no budget. |
| Exit cost | AROYA deletes data 90 days after termination and one 90,000 sq ft operator describes his exit as "a very long and difficult exit". Growlink's Terms carry no export obligation, and the manual CSV is a 20x downsample of what you paid to collect. |
The case against DIY
This section is deliberately unflattering to DIY. Someone should be able to read it and reasonably choose to buy. If the rest of this page is going to be credible about vendor costs, it has to be equally honest about what you give up by not paying them.
No support contract, and support is real value
Multiple independent first-hand reports value it highly: "their training, tech support, and software for the normal subscription is second to none. spend a year training, and in tech support with them twice a month, and you will benefit massively" (AROYA customer). "Their customer support team is incredible" (Growlink). At 3am with a room at 100% RH, there is nobody to call on a DIY rig.
No warranty, no RMA, no advance replacement
Growlink CORE includes 3-year advance replacement. TrolMaster warrants products 3 years. AROYA replaced a foliar-killed TEROS under warranty. Your DIY board dies and you buy another one and you rebuild it.
You are the integrator, and integration is where projects die
A Home Assistant SDI-12 thread runs to 26 posts over months: temperature-dependent moisture readings, the TEROS 12 soilless-media calibration polynomial topping out around 80% instead of 100%, unique-ID entity errors, serial parsing bugs. One participant: "the readings that get displayed don't match the readings from the little USB dongle thingy that you use with the Aroya app. So I reworked the equations and now I have a home brewed Aroya system." The calibration polynomial is a real hidden DIY cost.
Calibration is entirely on you, and it is harder than it looks
TEROS 12 is factory-calibrated for a reason. AROYA's own white paper reportedly concedes the TEROS 12 "wasn't accurate at high EC". Deriving your own media-specific VWC and EC transfer functions is a genuine soil-physics problem, not a coding problem.
There is no mature open-source alternative, only many small immature ones
As of August 2026 the leading crop-steering projects carry single and low-double-digit star counts. The leading project's maintainer posted in July 2026 describing "a serious silent failure: on some installs the engine couldn't find the zone moisture probes at all, so the zones froze... and watering fell back to a dumb timer." A user later the same month: "I can't get it to work... I've uninstalled everything. Too bad."
No compliance or audit trail
AROYA at $300/month reportedly includes METRC integration and the full platform. Argus Titan Enterprise at $28,000 buys multi-user, audit tracking and password control - explicitly a compliance product. A DIY stack gives you none of this, and in a licensed operation seed-to-sale integration and audit logging are not optional.
Silent failure risk is asymmetric
Growlink had an app bug that "will run your reservoir completely empty... and flood your tables" - but that was one company's bug, reported publicly, in a product with a support line. A DIY silent failure is yours alone to find, usually after the crop.
Key-person risk
A DIY system encodes its logic in one person's head. Commercial systems are documented, transferable, and survive that person leaving. Argus's "no user configuration or programming is required" is lock-in. It is also continuity.
Even DIY partisans concede the effort
An operator running Home Assistant with METER and Apogee sensors commercially: "No subscription cost, but a lot of coding and maintenance." And on the other side: "home assistant is way more effort than trolmaster and every update seems to break something."
The engineering gap is real
The fairest framing found in the research: "Are you going to design all the sensor front ends... Are you familiar with maintaining resolution of different bit depths of analog to digital converters? Are you going to design a PCB so its actually professional?" The conclusion drawn there: for microgrowers, off-the-shelf at $250-500 wins; for large operations, $4k software would be a no-brainer if it works.
What to actually buy
Different evidence class, read it that way. First-hand operator experience from running these products in a production facility. Not vendor-published, not third-party reported. Treat it as an informed opinion from someone with the kit in the ground, and weigh it accordingly.
Grow Sensor
Recommended without reservation
Genuinely excellent, and the one commercial product here that earns its money outright. It measures the things that are hard to measure well - PPFD, spectrum, DLI, leaf temperature and true leaf VPD - and it does it as a calibrated instrument rather than a lux sensor with a conversion factor bolted on. Getting real leaf VPD without building an MLX90640 node yourself is the headline: that is the number worth steering on, and most builds approximate it from air temperature because the direct measurement is awkward.
Why it fits a DIY-minded build: It integrates into Home Assistant, so it behaves like the rest of a DIY stack rather than a walled garden - the data lands next to everything else and you can build your own logic on top. There is no subscription to use the sensor or the app. Combined with the no-subscription position and a two-year warranty, it is the rare case where buying is simply better than building: replicating calibrated spectral measurement in DIY is genuinely hard, and this removes that problem for the price of a mid-range part.
Caveat. It is still a monitor. It measures the canopy, it does not run the room - pair it with a controller. And the forward-looking unknowns stand as the vendor states them: premium features "may require a subscription in the future", and GrowOps renewal pricing after the bundled first year is not published.
METER TEROS 12 substrate probes
The reference standard, and worth the premium
The AROYA-lineage substrate probes are the best in this catalogue and the ones to build around. Running both a TEROS 12 and a lower-cost SDI-12 clone in the same room, side by side, the TEROS 12 is clearly the better instrument - and substrate data is the one place where sensor quality directly changes irrigation decisions rather than just tidying a dashboard.
Why it fits a DIY-minded build: Crucially, buying the good probe does not mean buying the platform. The TEROS 12 speaks SDI-12 and DDI, METER explicitly document third-party logger use, and it costs a fraction of the AROYA subscription it is usually sold alongside. This is the clearest instance of the pattern the research found across sophisticated operators: buy the sensors, refuse the platform.
Caveat. Factory calibration is generic. The media-specific conversion still matters, and METER's own published figures make the point - plus or minus 0.03 m3/m3 on the generic calibration against plus or minus 0.01-0.02 with a medium-specific one.
Lower-cost SDI-12 substrate probes
Usable, and clearly second
The budget SDI-12 probes work and they are a legitimate way to get more zones covered for the money. Run against a TEROS 12 in the same substrate, they are the weaker instrument. If the budget only stretches to good probes in some zones, put the TEROS 12 where the irrigation decision is actually made and fill the rest with the cheaper ones.
Caveat. They need their own media-specific calibration curve, and the raw counts are not comparable to a TEROS 12's. Do not mix the two in one averaged zone figure.
Build recipes for the DIY side
The DIY figures on this page are not hypothetical - they come from working configurations in the parts catalogue, including a canopy multi-sensor node (CO², temperature, humidity, pressure, light, VPD, PPFD and DLI on one PoE node), a standalone CO² injection controller with a 60-second dead-man timer, and a full canopy stack that adds thermal leaf temperature and substrate sensing. Each carries the actual ESPHome config, scrubbed of credentials.
Buy the sensors, refuse the platform
The most consistent pattern across the user research is neither pure DIY nor pure platform. Operators at 30,000 and 90,000 sq ft run METER sensors into a PLC or pair AROYA with OpenSprinkler; others run METER sensors into ZENTRA Cloud at $180/yr instead of AROYA, or import Pulse data into Home Assistant over the documented API and write their own crop-steering logic. Pulse Grow is the natural bridge - documented self-service API, TEROS 12 retrofit kits at $39-69, and a working free tier.