How we assess AI ad tools
Every tool page on adtestbench rests on two kinds of evidence. The vendor’s own pages give the prices, credits, limits and terms, each read on the date shown. Public reviews on Trustpilot, Reddit, YouTube and G2 give what users report about output quality, support and billing, counted and linked. The verdict states what both show.
What each page rests on
Most of a page comes from documents you can open yourself: the vendor’s live pricing page, help centre, developer documentation and terms. Pros and cons come from public user reviews. Every claim sits on one rung of the evidence ladder below, and its wording shows which one.
| Rung | What it rests on, and how a page words it |
|---|---|
| Stated | The vendor’s own words, in quotation marks and attributed |
| Read | A primary document such as a pricing page or terms clause, cited with the clause number and the date read; an interpretation names the clause it rests on |
| Computed | Our arithmetic on sourced figures, shown with the formula and its inputs |
| Reported | Public user reviews and threads, counted against all those read for that tool and linked, never averaged into a rating |
| Inferred | Our judgement from the rungs above, with its reason |
Scroll sideways for all 2 columns
An example of each rung: editorial policy and corrections.
How prices are read and dated
A price is verified when it was read on the date shown, on the vendor’s live pricing page, official help centre or developer documentation. A price that could not be read there shows “Unverified”, and a vendor with no public price list shows “Not published”.
List price is a plan’s regular price on its official pricing page, before any coupon, partner deal or limited-time offer.
When this page was built, 95 of the 106 tools in the database had a verified price and 8 showed “Unverified”. Of the rest, 1 had no public price list and 2 had shut down.
Pages also print a cost per output at list price. It divides a plan, top-up or API price by what that price buys, then multiplies by what one output uses. Each tool’s method note shows its own sum. Where a plan counts credits, the figure assumes every credit is used and nothing is re-rendered.
How terms are quoted
A terms summary counts as verified only when the clause itself was read in the vendor’s terms. A plan card does not count. Each rights line quotes the clause with its number and the date of the version read, and ambiguous text reads “unclear”.
The AI commercial use matrix rates each plan and use from those clauses, with the date each was read.
How user reviews are counted
Each tool’s pros and cons come from one file of public reviews and threads. On 28 September 2026 those files held 1,972 reviews and threads across 52 tools. By site: 837 on Trustpilot, 473 on Reddit, 213 on YouTube, 139 in app stores, 129 on G2 and 87 on Capterra, plus Product Hunt, the Shopify App Store and independent blogs.
A point users raise prints with its count against the tool’s total, as in “18 of 64 sources” on Arcads, and links two of them. A point only one source raises stays off the page. Quotes run to 15 words at most.
A tool page’s source line gives the number of reviews behind it and the date the vendor’s pages were checked.
How often prices and terms are re-checked
Prices are re-checked against the live official page every day, and terms at least every 90 days. After 30 days a price stops counting as verified. Tables then replace the figure with “Unverified, last checked” and the date; sentences keep the figure beside the date it was last read.
Each re-check moves the checked date in a page’s meta line; “Updated” moves only when a fact changes, which the page’s list of changes records. The price tracker dates every tool’s prices.
Tools whose terms bar benchmarks
The terms below bar benchmarking the service or publishing the results. Meta’s go further and bar any use of its output off its own platforms. An agency that compares these tools side by side for a client, or publishes the comparison, reads these clauses first. The benchmark-ban register holds every tool’s clause, re-read on 12 October 2026.
| Tool, clause and date | The words that bar it |
|---|---|
| HeyGen Terms s.2, 15 Oct 2026 | “Access any portion of the services for benchmarking, comparative or competitive purposes” |
| Luma Terms 3.4(g), 14 May 2026 | “publish benchmarks or performance information about the Services” |
| AdCreative.ai Subscriber terms 2.2(f), 19 Dec 2025 | “use the Services for any competitive or benchmark purposes” |
| Pencil Core terms 5.1(i), 22 Apr 2026 | “engage in any competitive analysis or benchmarking of the Pencil Technology” |
| Claid.ai Terms s.3, item e, 10 Aug 2023 | “for benchmarking or competitive analysis of our Services” |
| Motion Terms, preamble, 12 May 2026 | “FOR ANY OTHER BENCHMARKING OR COMPETITIVE PURPOSES” |
| The Brief Acceptable Use Policy, “Engage in Competitive Analysis”, 1 Oct 2025 | “Use the Services for performance benchmarking or to build a competing product or service” |
| Canva Terms of Use 2(e)(iii), 19 Aug 2026 | “access the Service for purposes of performance benchmarking” |
| Jasper Usage Policies v1.2, A(iv), 24 Aug 2024 | “monitor the Services for any benchmarking or competitive purpose” |
| Meta Advantage+ creative Generative AI Terms s.1, 6 May 2024 | “Use or publication of Output outside of Meta’s platforms is unauthorized and a violation of these Terms” |
Scroll sideways for all 2 columns
Canva’s clause comes from the Wayback Machine capture of 23 September 2026, with the same effective date.
Jasper’s clause does not apply to a use expressly permitted, for example in its Documentation or an Order Form.
Meta’s clause names no benchmark. It bars publishing any Advantage+ creative output off Meta’s platforms. Its words come from the Wayback capture of 22 November 2024; the live page still showed the 6 May 2024 date on 24 September 2026.
Vidu’s clause is unclear. Its Terms of Service, section 5, last updated 3 July 2026, name no benchmark but bar “Using the Services for competitive analysis, developing competing products or services, or any purpose that may be detrimental to our business interests”. The terms do not say whether a published side-by-side comparison counts as competitive analysis.
Which tools are in the database
A tool enters the database when it makes ad creatives, edits them or helps test them. That covers AI UGC and avatar video, generated video, product video, static ads, ad copy and creative research.
A tool that shuts down or leaves ads keeps its one page, with no child pages or comparisons.
The ad creative brief template sets out every field, one by one.
Sources
- HeyGen: HeyGen Terms of Service, section 2, benchmarking bar, last updated 23 July 2026. Checked
- Luma AI: Luma Terms of Service, section 3.4(g), benchmarks and performance information, last updated 14 May 2026. Checked
- AdCreative.ai: AdCreative.ai terms of service for online subscribers, clause 2.2(f), competitive or benchmark purposes, updated 19 December 2025. Checked
- Pencil: Pencil core terms and conditions, clause 5.1(i), competitive analysis or benchmarking, updated 22 April 2026. Checked
- Claid.ai: Claid.ai Terms of Service, section 3, item e, benchmarking or competitive analysis, effective 10 August 2023. Checked
- Motion: Motion Terms of Service, preamble, benchmarking or competitive purposes, last updated 12 May 2026. Checked
- The Brief: The Brief Acceptable Use Policy, Engage in Competitive Analysis, performance benchmarking, last updated 1 October 2025. Checked
- Canva: Canva Terms of Use, section 2(e)(iii), performance benchmarking, effective 19 August 2026; clause read in the Wayback Machine capture of 23 September 2026. Checked
- Jasper: Jasper Usage Policies, version 1.2, A Platform Guidelines, item (iv), benchmarking or competitive purpose, effective 24 August 2024. Checked
- Meta: Meta Ad Creative Generative AI Terms, section 1, Rights in Ads Content, output outside Meta’s platforms, effective 6 May 2024; English wording from the Wayback Machine capture of 22 November 2024, and the live page still showed the 6 May 2024 date on 24 September 2026. Checked
- Vidu: Vidu Terms of Service, section 5, Prohibited Conduct and Content, competitive analysis, last updated 3 July 2026; the terms do not say whether a published comparison counts. Checked
What changed on this page
- MakeUGC’s terms of 22 October 2026 bar using its Services or Output to benchmark a competing product or AI model (7(f)).
- Rewritten to describe how reviews are made: official pricing pages, terms quoted by clause, and public user reviews counted and linked.
- Corrected: a price read in the vendor’s developer documentation also counts as verified (was: the pricing page or help centre only).
Alexandre S.
Alexandre S. is a creative ad strategist and copywriter with ten years' experience in digital marketing across multiple channels.
Related pages
-
About Editorial policy and correctionsHow every adtestbench page is sourced and dated, how user reviews are counted, how facts are re-checked, and how a correction shows on the page.
-
Data AI ad tool price tracker69 of 82 AI ad tools show a list price read on the vendor’s own page, last checked 16 October 2026, with the cost per clip or image where a rate allows.
-
Data AI commercial use, tool by tool and plan by planCan a tool’s output run in your paid ads? Terms quoted with clause and date for each plan: allowed, conditional, barred or unclear. Checked 26 September 2026.
