NACS Adapter: Which One Your Car Needs

The adapter in the frunk is the wrong one. Not defective. Pointed the wrong way.

That single fact is most of what changed for non-Tesla drivers, and it is why "can I use Tesla stations now" has four different answers depending on which piece of plastic you happen to own. The connector question settled at the standards level. The hardware bolted to concrete did not settle with it, and it will not for years: a pedestal installed in 2023 has its cable permanently attached, and the federal rule behind a lot of that concrete requires the charger to be maintained in compliance for not less than five years from its first day of operation (23 CFR 680.106(i)). Nobody swaps out a working cable on that schedule.

So treat this as a sorting exercise, not a news story. Two shapes exist on cars. Two shapes exist on cables. What you need is decided by which pair you are standing between, and the federal rules covering publicly funded stations still name one of those shapes and not the other, which is the part almost every summary has backwards.

Two shapes on cars, two on cables, four ways to be stuck

Both shapes are easy to tell apart once you have seen them. CCS1 is the round J1772 head with a second lobe underneath holding two large DC pins; the Joint Office of Energy and Transportation describes it as the J1772 five-pin layout "with two additional bottom 'DC' pins for use with a DCFC." J3400, the standardized form of NACS, is the small oval, and the same page notes it "works for AC (L1, L2) and DC charging utilizing the same pins for AC and DC power transfer" (SAE J3400 Charging Connector, read 28 August 2026). One connector doing both jobs is the actual engineering change. Everything below is the consequence.

Four directions, not one. A driver who has "the NACS adapter" has solved exactly one row of that table, and people routinely buy the row they already had covered.

That last column is deliberately soft. What arrives in the box is a per-model, per-model-year decision that moves between announcement and delivery, and the Joint Office puts it no more firmly than that manufacturers "have indicated that they will provide J3400 adapters to owners of CCS vehicles beginning in 2024." Treat any general claim about what your car came with, this one included, as a prompt to go and look.

The plainest definition of the object itself is in the same glossary: an adapter is "hardware that allows an EV to connect to a charger/station that it otherwise could not connect to directly." Notice what that sentence does not promise. Not that the session starts, not that it bills correctly, not that it runs at full power. Only that the metal meets.

The CHAdeMO row is not an oversight. Part 680 does let a federally funded site add one — "permanently attached CHAdeMO connectors can be provided using only FY2022 NEVI Funds" — and a single fiscal year of money with no continuing obligation is exactly as much future as that sounds like.

One piece of vocabulary saves real confusion at a busy site, and the regulation defines it for you. A charging port is "the system within a charger that charges one EV," and section 680.104 adds the sentence everyone skips: "A charging port may have multiple connectors, but it can provide power to charge only one EV through one connector at a time." Two cables hanging off one unit may be two ports, or one port offering a choice. If it is the latter, that second cable is not a second stall.

The federal rule still names CCS1, not J3400

This is where the popular version of the story goes wrong. Native-NACS cars are arriving, so the assumption is that federally funded stations are being built to match. Read the regulation and the opposite is true.

23 CFR 680.106(c), pulled through the eCFR on 28 August 2026 against the Title 23 issue dated 26 August 2026, reads:

Each DCFC charging port must be capable of charging any CCS-compliant vehicle and each DCFC charging port must have at least one permanently attached CCS Type 1 connector. ... Each AC Level 2 charging port must have a permanently attached J1772 connector and must charge any J1772-compliant vehicle.

No mention of J3400 anywhere in Part 680. None of NACS. None of adapters. All three words return zero hits across the ten sections, 680.100 through 680.118, in the version pulled that day. "Connector" is defined at 680.104 as "the device that attaches an EV to a charging port in order to transfer electricity," and the connectors the rule goes on to name are the two older ones.

Turn that around and the practical result is worth a note in the glovebox. If your car has a native J3400 port, a station built with federal highway money guarantees you a connector you cannot use without an adapter. CCS1 is the certainty there. NACS is the maybe. That is the reverse of what a lot of 2026 buyers assume when they hear the standard has switched.

There is a second number in that section that matters more for newer cars than any of this. Section 680.106(d)(1) requires DCFC ports to support output voltages between 250 and 920 volts DC. If you drive one of the 800-volt-class cars, that range is the reason a federally funded site is a safer bet for full-speed charging than an older private one, adapter or no adapter. The AC side carries its own floor one paragraph down, at (d)(2): at least 6 kW continuous per Level 2 port, and 6 kW per port simultaneously across every AC port at the station. Neither figure has anything to do with the shape on the end of the cable. A connector that fits tells you nothing about the voltage behind it.

The adapter can also live on the pedestal

Your car is not the only place this gets solved. Since a Federal Highway Administration issuance in May 2023, recorded on the same Joint Office page, a federally funded DC fast charger may carry a J3400/NACS adapter of its own, provided the CCS1 connector required by 680.106(c) is there as well.

The conditions attached are the part worth carrying. From that page: "For federally funded DCFC, adapters should be a permanently attached connector and have been integrated by the charger manufacturer," and the station "must meet all 23 CFR 680 standards and requirements, including but not limited to interoperability, power level, minimum uptime, and certification by a national recognized testing laboratory."

Read that as a quality floor rather than as red tape. A site-fitted adapter is not a loose dongle in a cabinet. It is factory-integrated hardware sitting inside the same obligations as the rest of the pedestal, including the average annual uptime of greater than 97 percent that 680.116(b) demands of every port. Tesla's Magic Dock, which the same page describes as "an attached NACS-to-CCS1 connector that allows a non-Tesla EV owner to plug in at a Tesla Supercharger station," is the version of this most drivers have heard of.

What you cannot do is tell from the road which sites have it. Which puts the pedestal-side adapter in the same bucket as everything else here: something to read off the individual station's page before you plan around it.

The adapter standard on the books covers the AC one

I went looking for the safety document behind these things expecting to find the DC adapter in it, because that is the adapter everyone argues about. It is not there yet.

SAE J3400/1_202504, Electric Vehicle Charging Adapter Safety and OEM-Qualified Device Designation, was issued on 3 April 2025 as an Information Report. Its scope covers "the general physical, electrical, functional, and performance requirements for adapters," states that the focus "is on defining the process to evaluate the suitability of adapters for NACS couplers," and then closes with the sentence that reframes the whole aftermarket: "This edition only covers adapters used between SAE J3400 and SAE J1772." That is the AC case. The revisions panel on that record listed no later edition when I read it on 28 August 2026.

For context, the family around it. J3400_202409 is the Recommended Practice for the coupler itself, revised 30 September 2024 and originally issued in December 2023. It covers "conductive power transfer to an electric vehicle using a coupler, which can be hand-mated and is capable of transferring either DC or AC single-phase power using two current-carrying contacts." Two contacts doing both jobs, in one clause. Alongside it, J3400/2, Connectors and Inlets for the North American Charging System (NACS) for Electric Vehicles, holds the dimensional definition of the connector and inlet; it was issued 3 April 2025 and its current revision, J3400/2_202505, is dated 27 May 2025.

Certification of the adapters themselves sits in a different building again. The Joint Office's published timeline names UL 2251 for connector and EVSE system testing and UL 2252 as the draft standard for adapters, with an industry consensus period opening across late 2023 and early 2024, and certified J3400-CCS adapters "anticipated to become available in the market" during 2024. Hold on to the tense there. A published timeline is a plan rather than a receipt, and that same page was still calling J3400 "in development from NACS" long after the Recommended Practice had been issued.

So the useful phrase to carry is the one in the title of that first document: OEM-qualified. Not "SAE compliant," which a product listing can print without meaning much, and not "UL listed," which describes a part rather than a pairing. Ask your automaker, in writing, whether a specific adapter is qualified for DC use on your VIN, and keep the reply. That email is worth more than any product page, and it is the same instinct that makes a permit worth filing away long after the inspector has gone.

A connector is the smaller half of what has to line up

Metal is the visible half. The other half is a conversation, and Part 680 specifies that one too. Section 680.108(a) requires chargers to conform to ISO 15118-3 and to carry hardware capable of implementing both ISO 15118-2 and ISO 15118-20, with charger software required to conform to ISO 15118-2 and to be capable of Plug and Charge by 28 February 2024. Paragraph (b) puts the charger on OCPP 1.6J or higher, then OCPP 2.0.1 by that same date. Paragraph (c) puts the networks on OCPI 2.2.1 for talking to one another.

None of that arrives through an adapter as a guarantee. The pins carry the handshake and the standards define it, but whether one particular car, adapter and station complete a Plug and Charge session together is an empirical question with one honest answer: test it once at a station near home, where a failure costs fifteen minutes instead of an evening. Carry a payment card that works on that network regardless.

An adapter never raises a ceiling

Two ceilings sit above every charging session, and no adapter moves either one.

On AC, the limit is your car's onboard charger, in kilowatts. A NACS-to-J1772 adapter on a Tesla wall unit does not turn a 7.7 kW car into an 11 kW car, for exactly the reason a bigger breaker does not: the car sets the rate, not the equipment.

On DC, the limits are your pack's acceptance rate and the station's voltage window. This one has a billing consequence. Where a site charges by the minute, a session slowed by cold, by a high state of charge, or by a mismatched voltage range costs real money per kilowatt-hour delivered, which is the trap laid out in per-kWh, per-minute, session and idle fees. An adapter that fits is not the same thing as a station that suits you.

And a bad connection can lower a ceiling. A DC adapter that gets hot enough to throttle a session is telling you something. The right response is to stop, not to re-seat it and hope.

Where a station's connector list is actually published

The rule that makes app data trustworthy is section 680.116(c)(8), which requires states and other direct recipients to make port-level fields available free of charge to third-party software developers via API, including, at (iii), connector types available by port. The neighbours in that list are worth knowing too: (v) power delivery rating in kilowatts by port, (vi) whether a port can be reached by a vehicle with a trailer, and (vii) real-time status by port "in terms defined by Open Charge Point Interface 2.2.1." The connector list on a federally funded site's app entry is a reporting obligation rather than a sales claim, and that difference is the whole reason to trust it.

Failures get counted somewhere as well. Under 680.112(a), states report to FHWA every quarter on each port's session start and end times, energy dispensed, peak session power, outage minutes and "any error codes associated with an unsuccessful charging session." None of that is a public per-station feed, and what does get released is aggregated and anonymised. Still, a bad connector at a funded site ends up in a file rather than only in a forum thread.

Three places to check, in order, before a trip you cannot improvise:

1. The network's own app, station detail screen. Connector types and power level per port.
2. The DOE Alternative Fuels Data Center station locator, which filters by connector type across networks and is neutral about brand.
3. Your automaker's charging page, for what is qualified on your car. These change often and they are brand-specific, so a general article cannot stand in for one. As a single dated example of what such a page actually says: Hyundai's US charging page states the company is "providing Tesla Supercharger-compatible NACS adapters for our current EV owners," that "starting with the 2025 IONIQ 5 and 2026 IONIQ 9, future electric models will come equipped with NACS ports standard," and that "Each 2025 IONIQ 5 will also come with a CCS adapter, so you can still charge at non-Tesla DC fast charging stations" (hyundaiusa.com, read 28 August 2026). One brand, one date, one model year. Yours will read differently, and that one will read differently again next spring.

Connector availability is binary in a way that fuel never was, which is why it belongs in route planning rather than in improvisation. Same argument as building real margins into a road trip, applied to plastic instead of miles.

Go and look at your charge port

Not the brochure. The car, in the driveway, tonight.

Open the charge door and note two things: whether the inlet is the two-lobe CCS1 or the small oval J3400, and what is physically in the vehicle with you right now. Count the adapters. Read the label on each one, and read it for direction rather than for brand, because an adapter is named for the two things it joins and half the confusion in this subject comes from reading that name backwards. Then find the row you are missing in the table above.

Most drivers turn out to own the adapter for the situation they were already fine in.