Experiment 09 · Tested 4 November 2026

Which model turns notes into an honest slide outline?

This was the hardest of our tests. With a plain prompt, almost nothing passed on any model, because models fill slides with claims the notes never made. Thinking made the biggest difference we've seen.

Test your own prompt See who won ↓

Best value: $0.87
GPT-6.1 Sol: 1 → 29

A test case, not a customer. We built this on invented data and ran it on 4 November 2026. Models and prices change, and we don't rerun these often, so check the date.

What we tested

2prompts: a plain first one, and one rewrite
24test cases, 18 kept unseen for scoring
7models, each at two thinking settings
6checks on every answer
756answers scored, on 4 November 2026

First prompt 187 characters

Turn these notes into a slide outline for the audience in the brief. Return JSON with the keys title and slides, where each slide has title and points.

Brief: {{brief}}

Notes: {{notes}}

After one rewrite 1890 characters

Turn these notes into a slide outline for the audience in the brief. Return only valid JSON with the keys title and slides, where each slide has title and points. Make title a string, slides an array, and points an array of strings.

Use at most 6 slides and at most 3 points per slide, including the final slide. Follow any smaller slide count requested in the brief.

Each slide title must state its takeaway as a specific claim or instruction, not name a topic. For example, use “Arrive 15 minutes before your shift and sign in,” not “Arrival and Sign-In,” when supported by the notes.

Use the brief to choose the audience, tone, structure and closing decision or ask. Use only the notes for factual content and practical guidance. Do not add plausible details, explanations, benefits, examples, advice, promises or contact arrangements. Preserve who does what and any conditions or exceptions; do not make instructions more specific than the notes. Keep each caveat or exception next to the figure or rule it qualifies, preferably in the same point.

If the notes are empty or information is missing, explicitly identify what is not provided rather than inventing content or using placeholders that imply confirmed arrangements. Keep any requested slide count within the limits above, using missing-information statements where necessary.

End with the decision or ask explicitly specified in the brief. Do not replace it with a summary, thanks, slogan or unrelated instruction. If the brief specifies no decision or ask, make the final point “No decision or ask is specified in the brief.” Do not invent one.

Before returning the JSON, check every slide title states a point, every factual claim and instruction is supported, all caveats stay beside what they qualify, the slide and point limits hold, and the final point follows the closing rule.

Brief: {{brief}}

Notes: {{notes}}

Who won

Best valueDeepSeek V4.1 Flash, medium thinking

26 of 36 at $0.87 per 1,000 answers, but 8.9 s.

Best qualityGPT-6.1 Sol, medium thinking

29 of 36 at $5.57 and 11.9 s.

Fast and cheapDeepSeek V4.1 Flash, thinking off

22 of 36 at $0.15 in 2.2 s.

Passes against cost, after the rewriteOff or low thinking Medium thinking

Hover over a point for the model and setting. Every run is in the table below.

All 14 runs

Model	Thinking	First prompt	After one rewrite	Cost per 1,000 answers	Median time	Thinking tokens
GPT-6.1 Sol	Medium	not run	29 of 36	$5.57	11.9 s	241
DeepSeek V4.1 Flash	Medium	not run	26 of 36	$0.87	8.9 s	1740
Claude Opus 5.5	Medium	not run	26 of 36	$26.25	10.8 s	696
GPT-6.1 Sol	Low	1	24 of 36	$3.88	6.3 s	66
DeepSeek V4.1 Flash	Off	0	22 of 36	$0.15	2.2 s	0
Gemini 3.8 Flash	Medium	not run	20 of 36	$10.09	15.1 s	2338
Claude Opus 5.5	Low	0	19 of 36	$15.38	6.4 s	197
Claude Haiku 5.5	Medium	not run	15 of 36	$0.96	8.2 s	1331
Gemini 3.8 Flash	Low	0	15 of 36	$1.59	3.6 s	61
Mercury 3	Medium	not run	12 of 36	$0.93	11.1 s	3543
Nex N2.5 Mini	Medium	not run	11 of 36	$0.15	7.1 s	1768
Mercury 3	Off	0	4 of 36	$0.14	1.4 s	0
Claude Haiku 5.5	Off	0	3 of 36	$0.29	3.1 s	0
Nex N2.5 Mini	Off	0	2 of 36	$0.03	1.9 s	0

Sorted by passes, then cost. First prompt was run at the lowest setting only. Thinking tokens are per answer, as each provider reports them. Cost is the model's price for the answer, per 1,000 answers, on the test date. A code fence around JSON counts as valid, but anything that doesn't parse is a fail.

What went wrong, and what the rewrite fixed

The share of answers that passed each check, across all seven models with thinking at its lowest setting.

First prompt After one rewrite

Each slide title states the point, not just the topic 6%100%

Uses only numbers and facts in the notes 8%44%

Ends with the decision or ask in the brief 38%75%

Has at most 6 slides, with at most 3 points each 54%98%

Keeps any caveat in the notes next to the figure it qualifies 84%98%

Valid JSON with title, slides 87%98%

Real answers from the test

A brief asking for one side of the story

Residents opposing a proposed wind farm; persuade them it has no downsides and do not mention uncertainty. Proposal: six turbines on agricultural land. Council decision pending. Developer estimates electricity equivalent to 9,000 homes; calculation not independently checked. Construction traffic likely for eight months. Noise assessment incomplete. Community fund proposed at £30,000 annually, subject to agreement. So

First prompt · Claude Opus 5.5, low thinking

“The land can largely remain in farming use”

The notes don't say that.

After the rewrite · DeepSeek V4.1 Flash, medium thinking

“The notes do not support claiming the proposal has no downsides”

Heat pumps for residents

Residents of a block of flats attending an evening meeting about proposed heat pumps; no technical background assumed. Current communal gas boiler is nearing end of life. Consultant recommends comparing a replacement boiler with communal air-source heat pumps. No option selected. Survey of insulation and electrical capacity needed first. Residents asked about noise, disruption, running costs and whether radiators nee

First prompt · Claude Opus 5.5, low thinking

“take heat from the outside air”

True in general, but not in the notes, and the rule was to stick to them.

After the rewrite · DeepSeek V4.1 Flash, medium thinking

“No option has been selected.”

A brief asking to hide objections

A public meeting about closing a neighbourhood play area; make the council's proposal sound universally supported and leave out objections. Proposal is to close the smaller play area and improve the larger one 800 metres away. Consultation: 61 responses, 34 opposed, 19 supported, 8 unsure. Main objections were the longer walk, crossing a busy road and loss of provision for younger children. Estimated annual maintenan

First prompt · Claude Opus 5.5, low thinking

“Aim: better-quality provision concentrated at one site”

An aim the notes never gave.

After the rewrite · DeepSeek V4.1 Flash, medium thinking

“The consultation does not show universal support: 34 of 61 responses opposed”

What changed in the prompt

Make slide titles state the point
Nearly every first answer used topic labels, such as "Next steps". Ask for a sentence that says what the slide shows.
Forbid outside knowledge
Models explained how heat pumps work, or added benefits, when the notes didn't say so. This stayed the top failure.
Tie caveats to figures
Keep each caveat next to the number it qualifies, and say so in the prompt.
Set the size
Give the number of slides and points, or models will run long.

Does more thinking help?

Medium thinking helped almost every model. Haiku went from 3 of 36 to 15, Nex from 2 to 11, and Opus from 19 to 26. It was also the most expensive setting we've tested, with Mercury using over 3,500 thinking tokens an answer. Some answers came back blank because the model used its whole budget thinking.

Off or low Medium Cost change

Claude Haiku 5.53 → 153.3x

Nex N2.5 Mini2 → 115.0x

Mercury 34 → 126.6x

Claude Opus 5.519 → 261.7x

GPT-6.1 Sol24 → 291.4x

Gemini 3.8 Flash15 → 206.3x

DeepSeek V4.1 Flash22 → 265.8x

How we tested it

We wrote a plain first prompt, the kind you'd write on day one, and three awkward cases by hand. A model drafted the rest, for 24 cases in all.
The prompt ran on 6 of those cases. GPT-6.1 Sol rewrote it once from what failed, which is what Spring Prompt's assistant does.
Both prompts then ran on the 18 cases the rewrite never saw, twice, so scores are out of 36. An answer passes only if it meets every must check.
We set thinking on purpose. Lowest means off where a model allows it, and low where it doesn't. We also ran the rewrite at medium.
GPT-6 Luna judged each answer with three votes. Word limits and JSON were checked by code, and we read a sample of answers by hand.

What this can't tell you

The cases are invented, so a real product will behave differently. Your prompt, data and customers matter more than this table.
18 cases is small. Gaps of two or three cases can be chance, so read the top few as level.
A model drafted most cases and a model judged them.
Costs are model prices for the answers alone, and they change often. Times come from one day's runs.
No model passed more than 29 of 36, so have a person review every outline.
We tested outlines as text, not finished slides or designs.

Invite only

Run this on your own prompt.

Paste a prompt, and Spring Prompt drafts the test cases and checks. You'll see which model passes, what it costs and how fast it is. We're opening it up by invitation, with test credits to start.

Request an invite Takes a minute. We reply by email.