1 · Manufacturer documents
Owner’s manuals, specification sheets, official model pages, warranty documents, and safety bulletins are the preferred sources.
Trust center
This is how we collect generator specifications, distinguish facts from calculations, screen for federal recalls, and decide when a product record needs another review.
754
catalog records
0
currently verified
39
CPSC generator recalls
September 16, 2026
recall snapshot
Source hierarchy
A product page should make it possible to follow important facts back to their source. We favor first-party technical documents and leave a field unknown when it cannot be verified responsibly.
Owner’s manuals, specification sheets, official model pages, warranty documents, and safety bulletins are the preferred sources.
Current manufacturer stores and authorized retailer listings help confirm availability, included equipment, and observed price.
Percentiles, medians, fit scores, and calculator results are derived from published fields using the disclosed rules below.
We do not invent missing noise, runtime, price, weight, capacity, or accessory compatibility values.
Offer ranking
A seller page must identify the exact model. Broad searches and category pages do not count as verified offers.
Lower current listed prices earn more points. A delivered total is used only when shipping is known.
Confirmed in-stock status ranks above low stock, preorder, unknown status, and out-of-stock records.
Free or flat-rate shipping can create a comparable delivered price. Calculated or unknown shipping stays open.
Published return windows and warranty terms improve confidence without overriding price or stock.
Recently checked price and availability data ranks above records that need another review.
Affiliate status and commission rate are excluded from the ranking. If only one verified seller exists, the site labels it as the best current destination—not as proof that it is the lowest price on the internet. Every offer also exposes a data-confidence score based on which price, stock, shipping, return, warranty, and freshness fields are documented.
Hands-on test results
A protocol, product sample, or unpublished worksheet is not a completed result. The public test ledger reads only records that have passed every gate.
The exact unit, instruments, conditions, valid and excluded runs, evidence files, and acquisition terms are complete.
The summary, comparisons, verdict, conflicts, and limitations accurately reflect the retained measurements.
Only then can the session appear publicly and place an FMG Tested badge on the exact catalog model.
Draft, pending, rejected, and withdrawn sessions never appear in the public API. Published results retain disclosed excluded runs and limitations instead of silently removing inconvenient observations. Browse the hands-on results ledger.
Freshness system
Technical specifications are reviewed on a 30-day target. The status ages automatically on the site so an old verification date cannot continue looking current.
Recall screening
Always verify directly. Before buying a used generator or operating an unfamiliar unit, check its complete model and serial number with the manufacturer and CPSC. If a recall might apply, follow the official stop-use and remedy instructions.
Owner reliability
Every submission begins as pending. Contact and proof details are never returned by the public review feed.
A report must pass ownership verification and content moderation before it can appear with a verified-owner label.
Averages and percentages stay hidden until five verified, published reports exist for the exact model.
Long-term figures use only reports with at least 12 months in service and require their own five-report sample.
Verified ownership establishes that the reviewer can reasonably document possession of the product; it does not turn an individual experience into a controlled test. Climate, loads, fuel, storage, maintenance, and usage patterns still vary. See the complete owner reliability standard.
Scoring and calculations
Products are compared with similar power type, design, and output bands. Missing fields are excluded and remaining weights are normalized.
Power, budget, fuel, portability, noise, and intended use are weighted against the visitor’s saved preferences.
Running loads are added, the largest motor-starting jump is included, and the chosen safety headroom is applied before rounding up.
Plain-language load packages set conservative running- and starting-watt targets. Matching models must clear both before fit, fuel, budget, portability, and sound preferences determine rank.
Each “Best generators for…” page applies fixed eligibility boundaries and page-specific weights to documented specifications, then diversifies the five displayed matches by brand.
Corrections are part of the system
Send the model, disputed field, and an official source link. We prioritize safety issues, recall information, and specifications that could change a sizing decision.