That won't change the required ~130mA coil current.
A buck converter could lower supply current, but not diodes or a resistor.
Leo..
2 diodes in series drop the voltage to around 1.4V
The 1N4000 series can handle up to 1A.
Arduino UNO can deliver up to 500mA on its 5V output.
Why couldn't it deliver 130 mA to the relays?
5 - 1.4 = 3.6, 3.6V / 25Ω = 144mA, 0.144 * 3.6 = 518mW into a 360mW coil = ye old magic smoke.
There will not be any smoke but you can add one more diode, lower the voltage to 2.9V (theoretical) and end of the problem.
Any other objections?
Regards
5 - 2.1 = 2.9V / 25Ω = 116mA. Coil power will be 0.116^2 * 25Ω = 336mW, but diodes will dissipate 0.116A * 2.1V = 244mW, 42% of the total 580mW.
That's OK if the OP don't care about wasting 1/4 watt. 
Two relays, one CO2 sensor, one oled screen, one Uno.
I count close to 450mA when the two relays are active.
Note that the drive circuit for that board must be 5volt, and power to the coil 3volt.
That involves removing a jumper, and using the diodes in the right place (across the jumper pins?). Get me the circuit diagram and I will tell you how.
Leo..
I do not find sense in the discussion when I only said to try with diodes as a second option (something that had already been said, by the way).
The first option is, and I repeat,
Regards
No offense intended, just pointing out some unforseen consequences
when trying to drop voltage with resistance or diode junctions. That "lost" power can sneak up and bite you. 
Regards
It is not the same to use diodes than a resistor.
Best regards
Well, voltage dropped * current = power dissipated, drive 2 amps through a 1N4004 diode and see how warm it gets, 2A * 0.7V = 1.4W of HEAT.
Maybe a better wording would be:
Cause 2 Amperes to flow through a 1N4004 diode.
If 2 Amps of DC current flows through a circuit and that circuit causes a voltage drop of 0.7 volts, then 0.7V * 2 Amps , 1.4 Watts of power will be dissipated by that circuit. Ohm's law.
But here we are talking about 2 coils that consume 130mA, nobody talked about a consumption of 2A more than you.
What is the point?
I think JCA is talking hypothetically, and I actually appreciate this type of discussion because I think it's where the most learning happens. And while I DO want it efficient, it's looking like I'll have to hop over to the project discussion section and start a new thread there outlining the whole project, cuz I realized literally last night that I can't make the lines to the CO2 sensor the length they need to be!! What a fool I've been. 
I have a multi-meter but never have used it to test current usage... do I use the leads in series, to test amperage? (I wouldn't be able to test the Uno, but I believe it's current use should be well-documented).
But even without testing, Leo mentioned the sensor and oled could be using about 150, 130 for relays, and 60ish for uno? I see posts saying 20 for uno, but regardless, Maximo, you said the uno can deliver 400/500mA on it's 5v rail... but 150+130+60 is still less than 400, which is why I was always a little confused. What numbers am I missing for your 450mA number, Leo?
That, and how do I make i2c lines work over a length of 3 feet >.<... Transistors? Ehhhh... I suppose I could try to relocate my power source area by my tent to be closer to where the sensor needs to be... or encase everything in expoxy 
I'm starting to feel like this is a raspbery pi sort of project.
Almost missed this. The circuit diagram for this particular relay? I could try... google lens-ing the chinese on the board might get me there...
What are you switching with those relays.
Leo..
The load on the AC side shouldn't matter since it's isolated? And well-within spec of the relay. At this point I'm only working with one relay, switching two different ac loads at the same time.
It's a 125W ultrasonic mister (think it's 24V) and a 12V computer fan (maybe noteworthy is I have that fan powered by a 19V wall brick, volt regulated to 12V with some ceramic caps, I could draw that schematic if you need?) All the relay is doing is closing an AC line, with a load of >3.5A, plugged into two female sockets soldered in parallel, with wires from a scrapped power cable... wires should be more than thick enough. The cable into the wall is one from the store, since I wasn't smart enough to think of the many computer PSU cables I have around. Wah wah -trombone sound-
But yeah, the fan blows into a tote with water and the mister, which goes up a pvc tube four feet up so the vapor reaches the top of a 7 ft high grow tent. The exhaust fan I wanted the second relay to control can totally run on low for the time being... Mushrooms can't have too much fresh air.
But maybe all that aside, I may just have to bite the bullet and get some 5v relays you've been subtly hinting at (reviews are questionable on the ones you linked tho 
hm...), because these relays are failing in a way that the 'normally open' is closed even when there is zero power to the relay, arduino..... that is what I call a catastrophic failure, since who knows what could have been running endlessly. And other relays I've foolishly taken out of their packaging are testing strangely... not closing at all, even tho the lights on..... idk, they look nice, for Chinese. But after the N.O. line being closed for no ding dang reason I should've thrown these all in a fire.
How do I know 5v relays will work any better? 
Let me know if you need a ironed out schematic! ... I hope to have this ironed out by next week 
Relay contact welding happens when you are switching inductive/complex loads, and you did not use a snubber circuit across the relay contacts.
Leo..
A ultrasonic resonator is a resistive load, but I got you. I was wondering about transients but is this not what relays are made for? Apparently not. But How do light switches work without arc degradation?? Are there snubber circuits everywhere? 
Oof, is the rabbit hole of electronics is dizzying.
I'm looking at snubber circuits, will it be feasible to make one? Am I in over my head 
Doesn't it have a switching power supply (which is not e resistive load).
A snubber circuit can be bought on a circuit board.
Leo..