The Category Ledger

Executive summary

No confirmed change released, 9 products showing source movement, 9 evidence limits lifted

• MARKET OBSERVATIONNo confirmed change was released in August 2026. 3 released earlier remain in the standing Record and are set out below. A period with no release is a period in which nothing cleared review, which is not the same as a period in which nothing moved.
• MARKET OBSERVATIONMonitoring ran on 5 days in this period and recorded source movement on 9 of 96 products whose sources were answering. Movement routes to review; none of it is reported here as a change.
• RECORD DEVELOPMENT6 products' sources stopped answering reliably during the period. Those products are named in the evidence section, and their held state keeps its observation date rather than being described as current.
• RECORD DEVELOPMENT9 recorded evidence limits lifted this period, which means new evidence became available on a dimension the Record had recorded as unobservable. That is FormulaSignal learning more, not the market moving.
• MARKET OBSERVATIONAmong the 65 products whose current label lets caffeine be computed line by line, the median is 242 mg and the range runs 100 mg to 400 mg. 33 sit under 250 mg and 30 sit at 300 mg or above, with 2 between them.
• MARKET OBSERVATIONPrice per serving across the 89 products where the arithmetic describes one package runs $0.33 to $3.75, with a median of $1.50 and the middle half between $1.12 and $1.83. Every one is an observed price, and none of them records a list basis.
• MARKET OBSERVATIONAcross the 80 products disclosing an amount for every line, beta-alanine appears on 67 and creatine on 28. In this cohort a product not declaring an ingredient is established rather than assumed, because every line on those panels carries an amount.
• RECORD DEVELOPMENTFigures in this edition describe 101 covered products across 22 brands, read as of August 26, 2026. Coverage is a defined set, not a census of the category, and 44 products carry at least one recorded limit on what can currently be observed.

A market observation describes something the Record observed about a product or its commercial context. A record development describes a change in FormulaSignal’s own evidence. A figure that moved between editions for the second reason is not a category event.

What changed

A confirmed change enters an edition on the date FormulaSignal released it, not on the date the underlying states were observed. Both dates are printed on every entry.

No confirmed change was released in August 2026. A period with no release is a period in which nothing cleared review, which is not the same as a period in which nothing moved. What monitoring did observe is below, and the standing Record follows.

Category benchmarks

Caffeine

Lowest

100 mg

25th percentile

200 mg

Median

242 mg

75th percentile

305 mg

Highest

400 mg

Products

65

• 65 of 101Current label states a caffeine total that can be computed line by line
• 19 of 101Current label declares the product stimulant-free
• 2 of 101Label states a caffeine total without declaring the lines behind itThe total is on file and is not counted into the distribution, because the lines it came from cannot be checked.
• 15 of 101Caffeine not computable from the current labelA proprietary blend, an unconfirmed transcription, or no caffeine line on the current panel. Not a claim that the product contains none.

Serving economics

Lowest

$0.33

25th percentile

$1.12

Median

$1.50

75th percentile

$1.83

Highest

$3.75

Products

89

• 100 of 101Products with a dated observed price
• 89 of 101Products where price per serving is arithmetic over one package
• 0 of 100Price observations recorded on a list basis by the sourceEvery other observed price may have been promotional. FormulaSignal does not infer a list basis from the absence of a strike-through, and no figure in this edition is called a list price.

Excluded from price per serving: RAW Nutrition Thavage RTD Pre-Workout, because the container holds one serving, so an observed package price does not establish what it covered.

Formulation

Covered products whose newest first-party panel states an amount for every line it declares, with no proprietary blend withholding a split and no unconfirmed transcription. This is the only cohort where not finding an ingredient is evidence the product does not declare it.

7 further formulation dimensions are in the full edition.

Readings

Each reading names the fact under it, the benchmark that gives the fact meaning, and the interpretation that follows. A possible implication is stated as a possibility, because FormulaSignal observes the Record and does not forecast a market.

3 further readings are in the full edition.

Product comparison

Covered products at publishable depth that answer every current-state question and disclose an amount for every line on their current panel, ordered by coverage tier and then by brand. Not a ranking and not a recommendation. 6 of 74 eligible products are shown here.

A blank cell means the Record does not hold that value and never means zero. Every price is an observed price: no covered product's newest price observation records a list basis. A formulation cell reading “not declared” is a finding: every product in this table discloses an amount for every line on its current panel, so not finding an ingredient means the label does not state one.

A percentile is this product’s position inside the observed cohort named in the benchmark above, and nothing more. A positioning line is a percentile inside a named cohort of observed values, and nothing more. FormulaSignal holds no efficacy finding and publishes no verdict on whether an amount is adequate, so no product here is ranked, graded, or described as better or worse than another.

68 further products are in the full edition.

Historical depth

• 5 of 101Products holding an earlier state comparable to the panel on file today
• 8 of 101Products holding at least one comparable adjacent pair of dated formula statesA different question from the row above, and a larger answer: a pair of states that are both now historical is comparable to each other without being comparable to the current panel.

Historical depth is uneven and is never continuous. FormulaSignal holds dated states, not a running log: the gap between two states is a gap in observation, and nothing is inferred about what happened inside it.

Evidence and method

• Every current-state figure reads the newest confirmed first-party observation on or before this edition's data-as-of date. A page read today never establishes what a product declared last year.
• A change is released only after a person read the evidence and signed it, with a review date on the record. Automated monitoring may observe, acquire, compare and propose. It may not confirm.
• Observed prices carry the basis the source stated. No covered product's newest price observation records a list basis, so no figure in this edition is called a list or standard price and no percentage price change is published from two observed prices.
• Missing evidence stays missing. A product whose panel does not disclose an amount is counted as not disclosing it, never as declaring zero.

• 101 of 101Products holding a dated first-party current observation
• 89 of 101Products whose current first-party observation declares at least one ingredient line
• 96 of 101Products whose sources answered on the last check

452 tracked sources across 101 products. 67 recorded evidence limits across 44 products. An evidence limit is FormulaSignal's own record that a dimension cannot currently be observed, what was tried, and the condition that would end it. It is a stated boundary on the sample rather than a gap somebody forgot.

86 candidates are awaiting human review. Candidates awaiting human review. No candidate is named, described or counted toward any figure in this edition: an unreviewed candidate is not a finding.

Limitations

• Every figure describes 101 covered products across 22 brands. FormulaSignal does not claim to observe every product sold in the category, and nothing here is a market share, a sales figure or a shipment estimate.
• 21 of 101 products do not disclose an amount for every line on their current panel, through a proprietary blend, a line stating no amount, or no current first-party panel at all. Ingredient counts in this edition run over the 80 that do, because that is the only set where a product not declaring an ingredient means anything.
• No covered product's newest price observation records whether it was a list or a promotional price. Every price in this edition is an observed price on a stated date. No percentage price change is published, and nothing here is called a list price.
• 1 product is excluded from price per serving because the container holds one serving, so an observed package price does not establish whether it covered one unit or a case.
• Continuous monitoring records begin August 18, 2026, so this period is only partly covered by them. Counts of checking activity describe the part that is covered and are not a claim about the rest.
• 6 products' first-party sources refused, disappeared or redirected during the period. Their last known state keeps its observation date and is not described as current.
• Historical depth is uneven. 5 of 101 products hold an earlier state the Record will compare against the panel on file today. FormulaSignal holds dated states, never a continuous log, and infers nothing about the gap between two of them.
• An observation date is the date FormulaSignal read a source. It is not a manufacture date, a rollout date, or the date a product actually changed.
• 86 candidates are awaiting human review. None of them appears in this edition in any form, because an unreviewed candidate is not a finding.

Corrections

A correction to a finding travels with the finding. A frozen edition is never rewritten: the correction record is published beside the finding it touches, so a reader can tell what an earlier edition said from what the Record holds now.

• AMENDED Cellucor C4 Original, accepted August 3, 2026. List price $29.99 to $34.99 on the homepage signal card, with observed standard price and price per serving increased 16.7% as commercial context. Both endpoint offer observations for 2024-01-25 and 2024-02-26 carry price_basis UNKNOWN. The record does not establish that either archived price was a list price rather than a promotional one, so neither the List price label nor a percentage move between them was supported..
• AMENDED Beyond Raw LIT V2 Pre-Workout, accepted July 29, 2026. The V1 to V2 change window ended 2025-09-12. Neither endpoint capture was observed on 2025-09-12. The captures on file were observed 2025-08-01 and 2026-07-23, so the stated window end was a rollout date from secondary coverage rather than an observation.