Independent capacity comparison
What 1–2 kWh power stations actually deliver
Battery nameplate capacity is not the same as energy delivered through an AC outlet. A current same-method test puts four major 1 kWh brands within a narrow range—and shows why the outage load matters more than a one-point efficiency difference.
Result first: no runaway efficiency winner
The four current 1 kWh-class models in OutdoorGearLab's July 2026 table delivered 85%–87% of claimed capacity through AC. That two-point spread is smaller than the change created by a refrigerator's duty cycle, a different test load, reserve settings, or conversion method.
This page has no paid links, no active merchant relationship, and no brand-supplied review unit. It compares published evidence and applies the same Backup Power Finder planning method to standardized capacity classes.
Product references
Identify the stations behind the capacity numbers
These are the exact power stations in the shared-method comparison. Pairing one with compatible solar panels creates a solar-generator system; the station by itself is only the storage-and-inverter side of that setup.
These publication-safe identity cards link each model to its official product page without redistributing manufacturer product imagery. An official page is a specification reference, not our test photo or endorsement.
Current models under one test method
| Exact model | Claimed capacity | Measured AC output | Usable |
|---|---|---|---|
| Anker SOLIX C1000 Gen 2 | 1,024 Wh | 890 Wh | 87% |
| EcoFlow DELTA 3 | 1,024 Wh | 870 Wh | 85% |
| Jackery Explorer 1000 v2 | 1,070 Wh | 910 Wh | 85% |
| BLUETTI Elite 100 V2 | 1,024 Wh | 890 Wh | 87% |
| Jackery Explorer 2000 v2 | 2,042 Wh | 1,710 Wh | 84% |
Source: OutdoorGearLab's head-to-head capacity table, updated July 23, 2026. “Measured AC output” is energy delivered through the station's inverter in that test—not a universal efficiency guarantee. Continuous and startup output are separate constraints that must be verified for the exact model.
Why one model can produce two correct results
The original-generation Anker SOLIX C1000 makes the methodology problem visible. OutdoorGearLab measured 897 Wh through AC, or 85% of its 1,056 Wh rating. ComputerBase measured 993 Wh at a continuous 500 W load and 954 Wh at a 2,000 W load. Those figures are not interchangeable verdicts: they came from different loads and test procedures.
| Exact model and source | Rated | Test condition disclosed | Measured AC energy | Use here |
|---|---|---|---|---|
| Original Anker C1000 · OutdoorGearLab | 1,056 Wh | In-house AC energy test | 897 Wh · 85% | Independent reference |
| Original Anker C1000 · ComputerBase | 1,056 Wh | 500 W continuous load | 993 Wh · 94% | Load-dependent upper result |
| Original Anker C1000 · ComputerBase | 1,056 Wh | 2,000 W continuous load | 954 Wh · 90.3% | High-load result |
| EcoFlow DELTA 2 · StorageReview | 1,024 Wh | 270 W AC load | 850 Wh · 83.3% | Different-model method example |
| BougeRV FORT 1000 · StorageReview | 1,120 Wh | 350 W AC load | 970 Wh · 86.6% | Different-model method example |
The honest conclusion is not that one laboratory is wrong. Inverter overhead, load level, temperature, station settings, cutoff behavior, measurement equipment, and test duration can all move the result. Cross-brand rankings should use one method; other tests provide context.
Run the same outage through three capacity classes
The calculator's default essentials schedule budgets a refrigerator at 120 W for eight total running hours, modem and Wi-Fi at 20 W for 24 hours, a 10 W lamp for six hours, and 12 W phone charging for three hours. That is 1,536 Wh per day, 162 W if everything runs together, and a conservative 1,042 W startup check.
The scenarios below are our calculations—not completed product discharge tests. Each starts at 100%, holds a 10% emergency reserve, and applies the same disclosed 15% conversion-and-variation allowance.
| Capacity class | Planning-usable energy | Battery-only support | 24-hour energy result | Output requirement |
|---|---|---|---|---|
| 1,024 Wh | 783 Wh | 12.2 hours | Short by 753 Wh | ≥162 W continuous and ≥1,042 W startup |
| 1,070 Wh | 819 Wh | 12.8 hours | Short by 717 Wh | Same load requirement |
| 2,042 Wh | 1,562 Wh | 24.4 hours | 26 Wh margin | Same load requirement |
For this entered schedule, a typical 1 kWh station is roughly a half-day battery-only plan. The 2 kWh class barely clears 24 hours, leaving little margin for a warmer refrigerator, additional devices, battery aging, or colder conditions. Change the entered loads before changing the product.
Solar changes the energy balance, not the evidence standard
A hypothetical 400 W portable array accepted at 400 W, with four peak-sun hours and the calculator's 70% field-yield assumption, adds an estimated 1,120 Wh per planning day. Against the 1,536 Wh schedule, the remaining average deficit is 416 Wh per day.
400 W × 4 peak-sun hours × 70% field yield = 1,120 Wh/day estimatedUnder that planning-day average, the standardized 1,024 Wh case clears a 24-hour target with a 367 Wh modeled margin. It does not guarantee sunlight arrives before the battery empties, that the station accepts 400 W, or that weather repeats. Verify voltage, current, connector, polarity, array limits, and the current manuals.
How to use this comparison
- Choose the outage duration and load schedule before choosing a brand.
- Use same-method tests for ranking; use other laboratories to understand sensitivity.
- Keep watt-hours, continuous watts, startup watts, and solar-input limits separate.
- Check the exact generation. C1000 and C1000 Gen 2—or 2000 Plus and 2000 v2—are not interchangeable models.
- Rehearse the exact station, appliance, settings, and cables before relying on the plan.
The useful buying decision
If a 1 kWh station already clears your measured load and outage target, efficiency differences of a few percentage points should not force an upgrade. If the target misses by hundreds of watt-hours, compare a larger battery, a disciplined load reduction, or compatible solar against that specific deficit.
Run your own outage load →Source register