Industrial Robotics Hub
comparison June 28, 2026 · Marcus Renner

Universal Robots vs FANUC CRX: Cobot Spec Comparison (2026)

UR wins on TCP speed -- the UR20 hits 5,000 mm/s versus the CRX-20iA/L's 1,500 mm/s -- but FANUC CRX holds IP67 across its entire cobot line while UR tops out at IP65.

Universal Robots vs FANUC CRX: Cobot Spec Comparison (2026)

Neither line wins outright. Universal Robots is faster at the top end and carries the deepest certified-integration ecosystem of any cobot brand; FANUC CRX runs IP67 on every arm in the series and posts tighter repeatability as payloads climb. The split is real and consistent across weight classes: in the mid range, for instance, the UR20 reaches 1,750 mm and hits 5,000 mm/s tool-center-point speed against the CRX-20iA/L’s 1,418 mm and 1,500 mm/s — a decisive UR advantage on cycle time. Flip to repeatability and sealing, and the CRX-20iA/L counters with ±0.04 mm against the UR20’s ±0.1 mm and IP67 against UR’s IP65. For a dry, high-throughput light-assembly cell, UR’s ecosystem and speed are the faster path to commissioning. For a wet machining cell or a food-line environment where coolant and washdown are facts of life, FANUC’s IP67 floor is worth more than any speed headline.

How this comparison is structured

Every figure below comes from the Industrial Robotics Hub database. No manufacturer marketing pages, no third-party robot databases. Where a figure is not in the database it is either omitted or shown as ”—”. The comparison pairs robots by payload class so you are looking at like against like. You can build any custom side-by-side at the compare engine, which pulls live specs for every robot in the database. If you are still deciding whether a collaborative arm is the right choice for your cell at all, the cobot vs. industrial robot guide covers that question first.

Entry class (~5-12 kg payload): UR5e and UR10e vs. CRX-10iA and CRX-10iA/L

The entry class is where both brands overlap most cleanly. FANUC offers two variants at 10 kg — a standard-reach and a long-reach arm. UR spans a wider payload range with the UR5e sitting lower and the UR10e carrying 12.5 kg.

SpecUR5eUR10eFANUC CRX-10iAFANUC CRX-10iA/L
Payload5 kg12.5 kg10 kg10 kg
Reach850 mm1,300 mm1,249 mm1,418 mm
Repeatability±0.03 mm±0.05 mm±0.04 mm±0.04 mm
IP rating (arm)IP54IP54IP67IP67
Max TCP speed1,000 mm/s1,000 mm/s1,000 mm/s

At the small end, the UR5e posts the tightest repeatability in this class at ±0.03 mm, which matters for high-precision light assembly. The UR10e carries 2.5 kg more payload than either FANUC 10 kg arm — relevant when a gripper eats into headroom. Both UR arms are IP54, however, while both FANUC arms carry IP67. That two-grade difference is the margin between an arm that handles coolant mist with a safety buffer and one that does not. For a dry light-assembly or kitting cell, UR’s repeatability edge and broader payload step are real advantages. For any cell with coolant, the IP67 floor on both CRX-10iA variants is not a marketing claim — it is the spec that keeps the arm out of a repair cycle in year two.

The CRX-10iA/L’s 1,418 mm reach is the longest in this class, 118 mm beyond the CRX-10iA and 118 mm beyond the UR10e. If the cell geometry demands it, that reach can be the deciding factor. We ranked the CRX-25iA first for wet tending and discussed the IP67 requirement in detail in the best cobots for machine tending 2026 analysis.

Mid class (~16-20 kg payload): UR16e and UR20 vs. CRX-20iA/L

This is the class where the key contrasts between the two lines are sharpest. Speed, repeatability, reach, and IP protection all split along brand lines with no single arm sweeping the comparison.

SpecUR16eUR20FANUC CRX-20iA/L
Payload16 kg20 kg20 kg
Reach900 mm1,750 mm1,418 mm
Repeatability±0.05 mm±0.1 mm±0.04 mm
IP rating (arm)IP54IP65IP67
Max TCP speed3,000 mm/s5,000 mm/s1,500 mm/s

The UR20 is the most extreme case of the speed-versus-precision-and-sealing trade-off. At 5,000 mm/s TCP it is the fastest arm in the mid class by a factor of more than three against the CRX-20iA/L’s 1,500 mm/s. It also reaches 332 mm further. Against that, its ±0.1 mm repeatability is the loosest in any class in this comparison, and IP65 — while a step up from IP54 — still falls one grade below the CRX-20iA/L’s IP67 for sustained flood-coolant exposure.

The UR16e is the compact option at 900 mm reach. Its 3,000 mm/s speed and IP54 rating make it a capable arm for a short-reach dry cell where throughput matters more than envelope. It does not compete with the CRX-20iA/L on IP or repeatability.

For a high-throughput dry assembly line where cycle time is the constraint, the UR20 is not a compromise — it is the right tool. For a CNC tending cell with flood coolant and a part that needs consistent seating, the CRX-20iA/L’s ±0.04 mm repeatability and IP67 protection are the better answer. The machine tending applications page covers the cell-level decision in more depth.

Heavy class (~30-50 kg payload): UR30 vs. CRX-25iA and CR-35iB

The heavy class is where FANUC’s cobot line extends furthest beyond UR’s reach — both in payload ceiling and in IP protection.

SpecUR30FANUC CRX-25iAFANUC CR-35iB
Payload35 kg30 kg50 kg
Reach1,300 mm1,889 mm1,813 mm
Repeatability±0.1 mm±0.04 mm±0.08 mm
IP rating (arm)IP65IP67IP67
Max TCP speed4,000 mm/s1,500 mm/s1,500 mm/s

The UR30 carries 35 kg — 5 kg more than the CRX-25iA — and does so with a 4,000 mm/s TCP speed and IP65 protection. Its reach is notably shorter at 1,300 mm versus the CRX-25iA’s 1,889 mm, which matters when the arm has to reach into a machine envelope and back out to a part buffer from a fixed base position. For a payload-heavy cell in a dry or mist environment where reach is not the constraint, the UR30’s combination of payload and speed is competitive.

The FANUC CRX-25iA counters with 589 mm more reach, ±0.04 mm repeatability (versus UR30’s ±0.1 mm), and IP67 protection. It was ranked first overall in the machine-tending cobot analysis precisely because that combination — reach, repeatability, and coolant survivability — covers the widest range of heavy-tending scenarios. For a foundry or a wet casting cell where the UR30’s IP65 is the borderline, the CRX-25iA is the safer spec.

The FANUC CR-35iB is in a category of its own for payload: 50 kg, still at IP67 and 1,813 mm reach. No UR arm in the current lineup matches 50 kg in a collaborative format. For heavy-casting tending or large-part assembly where the part plus tooling approaches or exceeds 35 kg, the CR-35iB is effectively the only cobot answer between these two brands.

Key contrasts across the full lines

Four contrasts define the choice at every payload level and are worth stating plainly.

IP rating. Every UR arm across the e-series and UR15/UR20/UR30 line sits at IP54 or IP65. The UR15, UR20, and UR30 all carry IP65. Every FANUC CRX and CR arm in this comparison carries IP67. IP67 versus IP65 is the difference between an arm rated for temporary submersion and one rated for water jets — a meaningful gap in flood-coolant machining, pharmaceutical washdown, or food processing. If the cell has a hose-down cycle, FANUC holds the rating floor.

TCP speed. UR is substantially faster at the top of its range. The UR15 and UR20 both hit 5,000 mm/s; the UR30 reaches 4,000 mm/s. FANUC CRX arms top out at 1,500 mm/s. Where cycle time is the binding constraint in a dry cell, UR’s speed advantage compounds across every pick-and-place or tend-and-return cycle on the shift.

Repeatability consistency. FANUC CRX runs a flat ±0.04 mm across its entire cobot line regardless of payload class. UR starts at ±0.03 mm on the small arms (UR3e, UR5e, UR7e) and widens to ±0.1 mm on the UR20 and UR30. For a high-payload UR arm doing precision fixture loading, the ±0.1 mm figure requires checking against the tolerance stack of the application. FANUC’s consistency up the payload scale is a genuine differentiator for tending applications where part seating is critical.

Payload ceiling. FANUC’s CR-35iB reaches 50 kg — the top of the cobot class in our database for either brand. UR’s ceiling is 35 kg on the UR30. For applications above 35 kg, FANUC is the only option between the two brands.

The UR+ ecosystem factor

The one structural advantage UR holds that does not appear on any spec line is the UR+ certified partner ecosystem of over 1,000 gripper, sensor, vision, and application-kit integrations built for the e-series platform. For a first-time cobot buyer, or for a shop running high-mix with frequent end-of-arm-tool changeovers, the ability to specify a pre-certified cobot application kit and have it commission in hours rather than days is a real economic variable. FANUC’s advantage is the mirror of this: its CRX arms share a programming environment and controller architecture with FANUC’s industrial articulated line, which means a shop already running FANUC industrial arms has an existing skill base that transfers to the CRX without retraining.

The Universal Robots brand page and FANUC brand page both list the full current catalog with filterable specs for every robot each brand fields in our database.

Use-case verdict

ScenarioVerdictReason
Flood-coolant CNC tendingFANUC CRXIP67 vs. IP65/54 across the line
High-throughput dry light assemblyUR5,000 mm/s TCP on UR20; ecosystem speed
Heavy casting / large chuck (>30 kg)FANUC CR-35iBOnly 50 kg cobot between the two brands
Precision fixture loading, high payloadFANUC CRX±0.04 mm flat vs. ±0.1 mm on UR20/UR30
First cobot, rich integration ecosystemURUR+ 1,000+ certified partner integrations
Shop already running FANUC industrialFANUC CRXShared controller and programming environment
Long-reach heavy tending (>1,500 mm)FANUC CRX-25iA1,889 mm vs. UR30’s 1,300 mm
Mist-only light cell, tight budget integrationURIP54 sufficient; faster commissioning
Food processing washdown lineFANUC CRXIP67 standard; neither brand reaches IP69K in this class

How to pick for your cell

Work through four questions in order.

Is the environment wet or dry? If the cell has flood coolant, a hose-down cycle, or a food-processing washdown requirement, set your minimum at IP67. In that scenario, FANUC CRX is the only option between these two brands at every payload class. The cobot types overview includes IP rating as a filterable field if you want to see how the full cobot market maps to that threshold.

What is the part weight including tooling, with headroom? Add the gripper weight to the part weight and target at least 20-25% rated payload above that sum to cover inertia at speed and the penalty of holding a part off the flange at the end of reach. If the combined figure approaches or exceeds 30 kg, the field narrows fast: FANUC CRX-25iA, FANUC CR-35iB, or UR30 — and IP rating likely decides which of those three fits.

Is cycle time the binding constraint in a dry cell? If throughput drives the economics and the environment is dry or light-mist, UR’s speed advantage on the UR16e, UR20, and UR30 is real and measurable. The UR20 at 5,000 mm/s TCP will run a shorter takt time on a pick-and-place or load-and-unload cycle than the CRX-20iA/L at 1,500 mm/s. In a wet cell, that speed figure is irrelevant because the IP floor already eliminates UR20 from the consideration.

Does existing infrastructure favor one ecosystem? A shop with established FANUC programming skills and a FANUC robot service contract has a path to CRX deployment that avoids retraining. A shop that is new to cobots and wants to draw from a large pool of pre-certified application kits has a faster commissioning path via UR+. Neither is a technical argument — they are operational ones — but they determine real cost and timeline.

The compare engine lets you place any two arms from the full database side by side on all of these specs simultaneously, including arms not in this post. If the class or environment you are speccing falls between the pairs above, or if a third brand needs to enter the comparison, that is the right place to run the numbers before committing to a platform.


All specifications sourced from the Industrial Robotics Hub database, current as of June 2026. Payload and reach are headline manufacturer figures; effective payload at full reach is lower and varies by model. IP ratings reflect manufacturer-published arm ratings as recorded in our database. Figures not in the database are omitted rather than estimated.

Compare these robots