Which 18650 Battery Brand Wins?
A head-to-head test of 12 18650 battery options with the measured results for each. See how they ranked and watch the full test video.
LG MJ1
Price shown in test: three dollars and 42 cents
Affiliate link. We may earn a commission at no extra cost to you.
The measured results
Every number below is read straight from the test. Scroll sideways to see all measurements. Products are listed in the order they finished.
| Product | Rated Capacity | Weight | Internal Resistance | Capacity Test 1000m A | Cold Temperature Capacity | Cost Per Mah Per Penny | High Drain Voltage Sag | Fan Runtime Showdown |
|---|---|---|---|---|---|---|---|---|
| 1LG MJ1three dollars and 42 cents | 3500 mAh, 10A, made in Korea | 46.8g and 46.8g (both samples) | 27 and 27.3 milliohms | measured 3404 and 3389 mAh vs 3500 mAh rated | close to or over 90 percent of rated capacity | 10.5 mAh per penny | 3.87V at 9A, 3.19V at its 10A rating (its own rated max) | 307 minutes, best of all 12 batteries tested |
| 2Samsung 35Efour dollars and 89 cents | 3500 mAh, 8A, made in Korea | 48.79g and 48.94g | 18.7 and 19.1 milliohms | measured 3276 and 3234 mAh vs 3500 mAh rated (93 percent of rating) | close to or over 90 percent of rated capacity | 6.7 mAh per penny | not tested at 10A since only rated for 8A | not tested |
| 3Panasonicseven dollars and 89 cents | 3500 mAh, 10A, made in Japan (Panasonic Sanyo), most expensive battery tested | 47.78g and 47.76g | 22.2 and 21.5 milliohms | measured 3450 and 3484 mAh vs 3500 mAh rated | close to or over 90 percent of rated capacity | not tested | 3.95V at 2A, 3.87V at 3A (rated 10A max) | transcript renders the time as "two hundred in ninety three" (likely 293 minutes, a split/garbled number); kept as reported, not corrected |
| 4Samsung 30Q (INR 30Q)four dollars and thirty four cents | 3000 mAh, 15A, made in Korea or Malaysia | 45.77g and 45.84g (transcript said "forty five point eight for", read as 45.84) | 11.8 and 12.1 milliohms | measured 2985 and 3002 mAh vs 3000 mAh rated (99.8 percent of rating) | close to or over 90 percent of rated capacity | 6.9 mAh per penny | not tested | 277 minutes |
| 5Sony Murata VTC6five dollars and sixty seven cents | 3000 mAh, 15A, made in Japan | 46.65g and 46.67g | 11.9 and 11.6 milliohms | measured 2969 and 2961 mAh vs 3000 mAh rated | close to or over 90 percent of rated capacity | not tested | not tested | 262 minutes |
| 6LG HE4$2.99 | 2500 mAh, 20A continuous, 4000mA max charge rate, made in Korea | 45.82g and 45.95g | 12.3 and 12.3 milliohms | measured 2524 and 2524 mAh, slightly exceeding its 2500 mAh rating | very little impact from cold (grouped with Samsung 25R as best cold performers) | 8.4 mAh per penny | not tested | 242 minutes |
| 7Samsung 25Rthree dollars and 42 cents | 2500 mAh, 20A, made in Korea or Malaysia | 43.9g and 43.92g | 11.7 and 11.7 milliohms | measured 2545 and 2528 mAh, exceeding its 2500 mAh rating | very little impact from cold (grouped with LG HE4 as best cold performers) | 7.4 mAh per penny (transcript said "7 point for", read as 7.4) | not tested | 228 minutes |
| 8Sony Murata VTC5Asix dollars and forty seven cents | 2500 mAh, 35A, made in Japan (highest amp rating of all cells tested) | 47.9g and 47.92g | 9.4 and 9.2 milliohms, lowest of all 12 cells tested | measured 2459 and 2425 mAh vs 2500 mAh rated | close to or over 90 percent of rated capacity | described as about the same value as the Chinese brands, no exact figure given | 3.5V at 10A, 3.27V at 15A (best of 5 brands rated 15A+), 3.04V at 20A (best of 3 brands rated 20A+) | 214 minutes |
| 9BXEthree dollars and 92 cents | 9800 mAh (claimed), made in China per narrator's research | 35.05g and 36.28g | 55 and 53.1 milliohms | measured 1194 and 1244 mAh vs 9800 mAh claimed rating (12 percent of claimed rating) | well below 25 percent of rated capacity | not tested | 2.82V at 4A | 104 minutes |
| 10Ultrafiretwo dollars and seventy five cents | 9800 mAh (claimed), least expensive battery tested | 35.36g and 35.88g | 48.1 and 31 milliohms (two samples differ notably; kept as reported, not averaged or corrected) | measured 1230 and 1163 mAh vs 9800 mAh claimed rating (12 percent of claimed rating) | well below 25 percent of rated capacity | not tested | 3.8V at 2A, 3.5V at 4A, 3.27V at 5A | 102 minutes, a few seconds longer than ETSAIR |
| 11ETSAIRtranscript reads "three dollars and sixty says", likely a garbled price (kept verbatim, not corrected) | 9999 mAh (claimed), made in China | 35.7g and 35.4g | 57.1 and 46.1 milliohms | measured 1299 and 1298 mAh vs 9999 mAh claimed rating (13 percent of claimed rating) | well below 25 percent of rated capacity | not tested | 3.27V at 4A, 2.59V at 6A | 102 minutes, stopped before Ultrafire |
| 12CASTNOO$3.75 | 3000 mAh, made in China | 36.08g and 35.61g | 71.5 and 76.3 milliohms, worst of all 12 cells tested | measured 718 and 651 mAh vs 3000 mAh rated | well below 25 percent of rated capacity | not tested | 1.98V at 4A, dropped below 1V soon after (first to be eliminated from the load test) | 69 minutes, first of all 12 to quit |
How it was tested
- 1000mA charge then discharge capacity test vs each brand's rated mAh
- internal resistance measurement (milliohms) at full charge
- cold temperature capacity retention at 6F / minus 14C
- cost per mAh per penny (value calculation)
- voltage sag under increasing amp load, 1A up to each cell's rated max
- fan-powering runtime showdown, all cells fully charged, longest to power a fan wins
Data notes and caveats
This video does not crown one overall winner. Its thesis is a myth test on a claim (do cheap Chinese-branded cells with inflated 9800-9999 mAh ratings really deliver 3-4x the capacity of name brands), not a single-victor showdown, so winner/runnerUp/budgetPick are left null per spec even though the closing narration clearly concludes Korean (LG, Samsung) and Japanese (Sony, Panasonic) cells are high quality while the four Chinese cells (Ultrafire, ETSAIR, CASTNOO, BXE) only delivered 12-24 percent of their claimed capacity. The fan-powering runtime showdown is the only test that ranks all 12 cells against each other in one ordered list, so products[] is ordered by that result (LG MJ1 best at 307 minutes, CASTNOO worst at 69 minutes); LG MJ1 also had the best cost-per-mAh figure (10.5 mAh/penny), making it the closest thing to an implied top pick, but no transcript sentence declares it the winner. Brand order/prices in the transcript map cleanly 1:1 onto the description's 12-brand Products Tested list (Ultrafire, LG HE4, LG MJ1, Samsung 25R, ETSAIR, CASTNOO, BXE, Samsung 30Q, Samsung 35E, Sony VTC6, Sony VTC5A, Panasonic in ascending price order), so brand resolution is high confidence. The high-drain voltage-sag load test (1A-10A+) groups results as 'the first six/seven/eight brands' without always naming which cells are in that group; only explicitly-named per-brand data points from that test were kept, the rest were omitted rather than guessed at group membership. Two numbers are garbled and were kept verbatim rather than corrected: ETSAIR's price ("three dollars and sixty says") and Panasonic's fan-showdown finish time ("two hundred in ninety three", likely 293 minutes).