Test log
Presentations.AI test data
Raw results for the Presentations.AI review. The verdict is in the review.
- Tested
- 5 November 2026
- Tester
- Miles Foster
- Account
- Free demo account, signed up by RankDude, nothing paid to the brand
- Tests
- 10
- Screenshots
- 9
- Method
- Published methodology
- Environment
- Google Chrome on desktop
Results
| ID | Test | Key measurement | Score |
|---|---|---|---|
| T01 | Brand & Theme Control | – | n/a |
| T02 | Export Fidelity | – | n/a |
| T03 | Structure from a Prompt | Structure points met: 19 of 21 | 4/5 |
| T04 | Slide Design | Layout faults: 10 in 30 slides | 3/5 |
| T05 | Content Accuracy | Invented figures: 3 on one slide | 2/5 |
| T06 | Use of a Supplied Document | Key points covered: 15 of 15 | 2/5 |
| T07 | Editing Control | Edits that worked: 9 of 12 | 3/5 |
| T08 | Charts & Images | Charts right: 2 of 3 | 2/5 |
| T09 | Speed | Water cycle deck: 3 min 7 s | 3/5 |
| T10 | Pricing & Value | Cheapest plan with export: $20 a month, billed yearly | 3/5 |
| Overall (in the review) | 3.4/5 | ||
Brand & Theme Control
n/aNot Tested- Method
- Add the logo, two hex colours and a font, import a .pptx template and check every slide.
- Remark
- Not graded; left out of the average.
Raw notes
Add the logo, two hex colours and a font, import a .pptx template and check every slide.
Not run: the Brand Kit item in the sidebar carries a Pro badge and the plan chooser lists Brand Kit under Pro. The Style panel does offer ready-made themes on the free plan, which I used in the editing test.
- Needs a paid plan.
Export Fidelity
n/aNot Tested- Method
- Export each deck to PowerPoint, Google Slides and PDF and compare slide by slide.
- Remark
- Not graded; left out of the average.
Raw notes
Export each deck to PowerPoint, Google Slides and PDF and compare slide by slide.
Not run: the Download button opened the plan chooser, which says "Every paid plan includes editable PowerPoint and PDF export with no watermark." The Share box also lists Export to PowerPoint.
- Needs a paid plan.
Structure from a Prompt
4/5Tested- Method
- Each of the three first drafts was checked against the seven-point list in the methodology: the requested slide count within one, a title slide, an agenda or opening, a logical order, one idea per slide, a closing ask and no repeated slide.
- Structure points met
- 19 of 21
- Water cycle slides asked, got
- 12, 10
- Remark
- Scale 4: 17 to 19 of 21 points.
Raw notes
Each of the three first drafts was checked against the seven-point list in the methodology: the requested slide count within one, a title slide, an agenda or opening, a logical order, one idea per slide, a closing ask and no repeated slide.
The pitch deck scored 7 of 7: ten slides, a title, a problem opening, a clear order, one idea a slide and a $1.5M ask on slide 10.
The quarterly review scored 6 of 7. Its agenda slide lists Intro, Problem, Solution, Product, Business model and Next steps, which is the outline of a startup pitch and not of this review, so I did not count it as an agenda.
The water cycle deck scored 6 of 7. I asked for 12 slides and got 10, which is two away.
- Each brief was generated once.
Slide Design
3/5Tested- Method
- All 30 slides of the three first drafts were captured at 2x and checked for text overflowing its box, overlapping elements, misalignment, mixed fonts, walls of text, blurry or unrelated images and left-over placeholder text.
- Layout faults
- 10 in 30 slides
- Faults per ten slides
- 3.3
- Remark
- Scale 3: 3 to 4.9 faults per ten slides.
Raw notes
All 30 slides of the three first drafts were captured at 2x and checked for text overflowing its box, overlapping elements, misalignment, mixed fonts, walls of text, blurry or unrelated images and left-over placeholder text.
I counted 10 faults in 30 slides, which is 3.3 per ten. In the quarterly review, slides 7, 9 and 10 set their text very small at the foot of tall empty panels. In the water cycle deck, slide 2 puts an up arrow beside every share of the world's water, slide 4 shows a piggy bank on a slide about plants, slide 7 has two "KEY LABEL" tags and slide 10 prints "Visual Insight: Key Graphic Highlight".
In the pitch deck, slide 2 has three tall boxes that are mostly empty, and slides 4 and 5 put a hotel lobby and a living room behind coffee market figures.
- Text contrast was read by eye, not measured with a checker.
Content Accuracy
2/5Tested- Method
- Every figure in the quarterly review was compared with my table, and the water cycle deck was checked against the USGS Water Science School table of where Earth's water is.
- Invented figures
- 3 on one slide
- Chart labels not in the table
- 4
- Minor water cycle errors
- 1
- Remark
- Scale 2: four to six errors, or any invented figure.
Raw notes
Every figure in the quarterly review was compared with my table, and the water cycle deck was checked against the USGS Water Science School table of where Earth's water is.
Revenue, orders, new customers, tickets, margin and satisfaction all matched my table. But slide 9 of the quarterly review is a next steps slide about a software plan: "Annual plan at a 20% discount", "First value in under five minutes" and "Reps engage accounts past 10 seats". None of that is in my brief, and I had said not to invent figures.
The bar chart on slide 6 shows 840, 666, 605 and 471 beside the regions. Those are Q2 and Q3 added together, and they are not in my table.
In the water cycle deck the split of 97, 2 and 1 per cent is a fair rounding of the USGS table (96.54 per cent oceans, 1.74 per cent ice), but the 1 per cent is labelled "Available fresh surface water". The USGS table puts most liquid fresh water underground, so I counted that as one minor error.
- I checked the water cycle facts against one USGS page, not two sources.
Use of a Supplied Document
2/5Tested- Method
- The six-page Harbor Light Coffee Roasters PDF was uploaded and a ten-slide pitch deck was requested.
- Key points covered
- 15 of 15
- Invented claims
- 1
- Remark
- Scale 2: four to seven covered, or anything invented. Fifteen covered with nothing invented would be a 5.
Raw notes
The six-page Harbor Light Coffee Roasters PDF was uploaded and a ten-slide pitch deck was requested. The 15 key points were listed before the test and each was checked.
All 15 key points were present and correct: the 2023 start in Portland, Maine, the two founders, the $1.5 million round, the 90-day-old supermarket bag, the 20 kg minimum, the 48-hour dispatch, the subscription sizes, the 2 kg no-contract cafe order, the $48 billion market growing about 8 per cent, 3,100 subscribers and 40 cafes, 3.5 per cent churn, $2.1 million revenue at 38 per cent margin, the 71 and 29 per cent split, the $410 value and $62 cost, and the 40, 35 and 25 per cent use of funds with the $5 million goal.
One claim is not in the document: slide 8 says the second roaster will "double batch capacity". The scale counts anything invented as a 2, so the grade is 2 even though no point was missed.
- One run on one document.
Editing Control
3/5Tested- Method
- Twelve fixed edits were tried on the quarterly review deck: reorder, change a layout, regenerate a slide, shorten a paragraph with the AI, insert a slide, edit chart data, replace an image, add notes, move an object, change the theme, undo three steps and leave a comment.
- Edits that worked
- 9 of 12
- Credits for the AI edits
- 25
- Remark
- Scale 3: 7 to 9 edits.
Raw notes
Twelve fixed edits were tried on the quarterly review deck: reorder, change a layout, regenerate a slide, shorten a paragraph with the AI, insert a slide, edit chart data, replace an image, add notes, move an object, change the theme, undo three steps and leave a comment.
Nine edits worked with normal controls. I dragged a thumbnail to reorder, picked another layout through Remix, shortened a slide with Make shorter, added a slide below with a one-line prompt, changed a chart value in the Edit Data sheet, typed speaker notes, dragged a block by its grip and applied the Rhyme theme to every slide in one click. Redesign slide (Remix) counted for regenerating a slide, because it re-laid the slide out but kept its words.
Three did not: the image picker opened and I chose a picture, but the title photo stayed the same; seven clicks on Undo did not take the theme off; and I found no way to leave a comment for a collaborator.
- Undo was tested once, after the theme change.
Charts & Images
2/5Tested- Method
- The quarterly review asked for a bar chart, a line chart and a KPI slide from the table.
- Charts right
- 2 of 3
- Charts you can edit as data
- line chart tested
- Remark
- Scale 2: two or more charts wrong, or images with no licence.
Raw notes
The quarterly review asked for a bar chart, a line chart and a KPI slide from the table. Each value was checked, each chart was clicked to see whether it is a real chart, and the images were checked for relevance and a stated source.
The line chart is right: July 412, August 455 and September 501 in $k. It is a native chart, with Change, Add Point and Edit chart buttons, and I changed a value in its data sheet. The KPI slide shows $1,368k, +12.7 per cent, 41 per cent and 4.6 out of 5, all as given.
The bar chart is wrong. It stacks Q2 and Q3 and labels each bar with the total, so North reads 840 where my table has 388 and 452.
None of the pictures shows where it comes from. The image picker and the Terms of Use say nothing about who owns the stock photos, and several are unrelated to their slide: a piggy bank, a hotel lobby, a living room.
- I did not click the bar chart to test its data sheet.
Speed
3/5Tested- Method
- Each brief was submitted from the create box, and the clock ran until the editor opened on the finished deck.
- Water cycle deck
- 3 min 7 s
- Quarterly review
- 4 min 28 s
- Credits per deck
- 100
- Remark
- Scale 3: median 2 to 3.9 minutes.
Raw notes
Each brief was submitted from the create box, and the clock ran until the editor opened on the finished deck. Between the two there is a question box and a thinking screen that streams its notes.
The water cycle deck took 3 minutes 7 seconds. The quarterly review took 4 minutes 28 seconds. Both stopped on a box that asks "What company is this presentation for?" and the deck only arrived after I closed it. The pitch deck sat on the same box for more than 20 minutes while I was away and finished within a minute of my closing it, so I did not count it.
The median of the two timed decks is 3.8 minutes. Each deck cost 100 credits.
- Two timed decks; the third stalled.
Pricing & Value
3/5Tested- Method
- Plans were read in the plan chooser inside the editor and on the public pricing page, which says its prices are for the US.
- Cheapest plan with export
- $20 a month, billed yearly
- Remark
- Scale 3: $15 to $24.99 a month for the cheapest plan that exports a .pptx file, billed yearly.
Raw notes
Plans were read in the plan chooser inside the editor and on the public pricing page, which says its prices are for the US.
The cheapest plan that exports a PowerPoint file is Pro at $20 a month, billed yearly, with 5,000 credits a year. Gold is $100 a month for 50,000 credits. The Starter plan is free; the pricing page says 200 credits and Up to 20 Slides, while my account showed 400 credits.
The plan chooser in the editor showed prices in shekels, because of my region, so I read dollars from the public page.
- No plan was bought, so I could not see a paid export.
Other features checked
Coverage
Tested
- The three fixed briefs, each sent once, with the pitch deck built from the uploaded six-page PDF
- Seven structure points, layout faults, every figure against the table and 15 key points of the PDF on all 30 slides
- Twelve fixed edits on the quarterly review deck, by hand and with the Edit with AI menu
- Download menu and plan chooser, public pricing page, Terms of Use and Trustpilot profile read
Not tested or limited
- Each brief was run once, so every grade rests on one deck per brief
- Brand control was not graded: the Brand Kit item in the sidebar carries a Pro badge, so the logo, colours, font and .pptx template could not be applied on the free plan
- Export fidelity was not graded: the Download button opened the plan chooser, so I never held a PowerPoint, PDF or Google Slides file
- Text contrast was read by eye, not measured with a checker
- Image replacement was not confirmed: the picker opened and I chose a picture, but the title photo did not visibly change
- The pitch deck stalled on a company website question while I was away, so its time is not counted
Still to test
- Brand Kit and template import on a paid month
- Export fidelity to PowerPoint, Google Slides and PDF on a paid month
- The image replacement and the undo button once more
- The brief with an immediate answer to the company question, to time all three decks the same way