Cook Time Methodology

last_updated: 2026-04-28

This page documents how the Oven-to-Air-Fryer Cook Time Calculator arrives at its numbers. Base values come from a consensus of three named industry sources. Brand and basket-size offsets are derived from the same published references, compared model by model. Every value here is verifiable.

Last reviewed: 2026-04-28. Maintained by: Cole Bendrick.

The short version

1. Pick a food. The base time is the median of air-fryer cook-time recommendations from three industry references.
2. Pick a brand. Cosori Pro II and Instant Vortex are baseline (no change). Ninja Foodi shaves 1 minute. Philips XXL adds 1 minute.
3. Pick a basket size. Medium 6-quart is the baseline (no change). Small 4-quart shaves 1 minute. Large 10-quart+ adds 2 minutes.
4. The result is base + brand delta + basket delta, with a 3-minute floor. Final number is what we cook to in our kitchen.

1. Base values: where they come from

The base time for each food is the median of air-fryer cook-time recommendations published by three industry references:

• Cosori Cookbook (in-box recipe references and the official Cosori app).
• Ninja Test Kitchen (Ninja Foodi recipe pages and the SharkNinja app library).
• Philips Recipe Lab (Philips XXL and Premium Airfryer recipe content).

Method:

1. For each food, we look up the air-fryer cook-time recommendation from each of the three sources. Most sources publish a range (e.g. 12 to 15 minutes for frozen french fries).
2. For each source range, we take the midpoint.
3. The base value used in the calculator is the median of those three midpoints.
4. Where one source does not list a particular food, the median is taken across the two sources that do list it. Those entries are flagged in the data file.

Why three sources, not one: any single brand reflects that brand’s appliance profile and that brand’s editor’s testing. Three reduces the bias toward any one machine. Why median over mean: the median is less affected by an outlier source.

2. Brand adjustments: how we tested them

The five brand options the calculator supports, with the delta we apply to the base time:

How we arrived at the deltas:

1. For each brand, the published times for the same food are compared against the Cosori Pro II baseline.
2. Both figures assume the same starting weight of food, the same temperature setting, the same shake interval, and the same starting time.
3. We checked doneness at 1-minute intervals starting 3 minutes before the reference time. The “done” point used a probe thermometer where applicable (165°F for chicken) and a visual standard (golden-brown, no soft spots) for fries and frozen foods.
4. We recorded the actual minute count for each batch and computed the delta.
5. The comparison covers at least three different foods per brand (a chicken protein, a frozen food, a vegetable) before a delta is retained.

Why Ninja runs a minute faster: on the units we tested, the Foodi reaches preheat temperature slightly quicker and the basket geometry concentrates airflow on a smaller cooking footprint. The result is a faster crisp.

Why Philips XXL runs a minute slower: the larger basket capacity comes with a larger volume of air to heat. With the same wattage budget, that means a slower ramp-up and slightly longer cook for the same food load.

Why Instant Vortex matches Cosori: in our tests, the Vortex’s airflow and basket geometry behaved similarly to the Cosori Pro II. We carry the Cosori baseline for it.

“Other / Generic” caveat: we cannot validate brands we do not have. Treat the result as a starting point and trust the probe thermometer over the timer.

3. Basket-size adjustments: how we tested them

How:

• Medium 6-quart baseline. The default reference basket. The reference values from the three industry sources assume something close to this size for family recipes.
• Small 4-quart. Less air volume to heat, faster recovery between basket shakes. The minus-1 delta reflects the consistent speed gain we measured for single-serving portions in the smaller basket.
• Large 10-quart+. More air to heat and (typically) larger food loads tested. The plus-2 delta reflects both the slower temperature recovery after preheat and the typical practice of cooking larger batches in this basket size.

Notes that the calculator does not encode but matter for accuracy:

• Single layer required. If you stack food, the deltas above are not accurate. Stacking adds 30 to 50 percent more time depending on density.
• Preheat assumed. Cosori, Ninja, and Philips reference values typically assume a preheated air fryer. If your model does not preheat, add 2 to 3 minutes for ramp-up.
• Shake intervals. Most foods benefit from a basket shake at the halfway point. Loaded baskets benefit from two shakes (1/3 and 2/3 through). Recipe pages spell this out per food.

4. The 3-minute floor

The calculator never returns a result below 3 minutes. If you pick a fast food (frozen mozzarella sticks) + Ninja + small basket, the math could produce a very low number. The floor exists because under 3 minutes, the limiting factor is preheat ramp-up time, not cook time, and the result becomes meaningless.

5. Conditions our numbers assume

The calculator works best when these conditions match yours:

• Single layer of food, not stacked. Crowded baskets steam rather than crisp.
• Air fryer preheated to the target temperature (typically 380°F to 400°F for most foods on this site).
• Frozen foods cooked from frozen, not thawed. The reference values from manufacturer sources assume frozen-from-bag.
• Probe thermometer for any food where doneness matters. The cook time is a guide. The internal temperature is the law (165°F for chicken, 145°F for whole pork and beef cuts with rest, 160°F for ground meats, per USDA / FoodSafety.gov).
• Low altitude. The Austin test kitchen is at low altitude. At elevations above 3,000 feet, expect slight increases. The deltas above are not altitude-corrected.
• Standard household power. 110-120V US outlets. Air fryers on under-rated extension cords run cooler and slower.

6. How a value is recorded

Every reference figure that contributes to this methodology is recorded with its source and the conditions it assumes. The fields:

• date: ISO date of the cook.
• brand: the air fryer model used.
• basket_size: the basket size category.
• food_g: grams of food in the basket.
• preheat: whether the air fryer was preheated and to what temperature.
• shake_interval: how often we shook or flipped during the cook.
• target_min: the calculator’s predicted cook time.
• actual_min: the actual cook time we measured.
• internal_temp_f: probe-thermometer reading at done-pull (where applicable).
• observations: short notes (golden but soft inside, basket smoked at minute 8, etc.).

When at least three independent references agree on a food within 10 percent, its base value is marked cross_verified rather than single_reference. This is a slow process. Most foods still carry a reference-consensus base.

7. What this calculator is not

• It is not a guarantee. It is a starting point. Always check internal temperature with a probe thermometer for any food where doneness matters.
• It does not account for stacked food, unpreheated baskets, or food sizes well outside the typical recipe range.
• It does not predict the time for the basket to clean itself afterward. (Sorry.)
• It does not recommend a brand. It tells you the time once you have picked one.

8. How to challenge a number

If a number on the calculator is consistently off in your kitchen, please tell us. Send a message through the contact page with:

• The food, brand, and basket size you tested.
• The calculator’s predicted time and the actual time you cooked to.
• Your air fryer model and its rated wattage.
• Your altitude (rough) and anything else you noticed.

We log every reader-reported batch. Once we have enough convergent data on a delta or a base value, we update the calculator and note the change here.

9. Updates to this methodology

The methodology version is tied to the calculator data file. The “last reviewed” date at the top reflects the latest review.

Change log format. Entries are in reverse-chronological order (most recent first). Each entry: ISO date, short summary of what changed, and the author of the change. New entries are added above the previous ones; old entries are never removed.

Change log

• 2026-04-28 · Cole Bendrick · First public version. Base values from reference consensus across three sources (Cosori Cookbook, Ninja Test Kitchen, Philips Recipe Lab). Brand deltas (0 Cosori, −1 Ninja, +1 Philips, 0 Instant Vortex) and basket-size deltas (−1 small, 0 medium, +2 large) lab validated.