Ranked by what each change is worth on the composite — the gap to the next rung of the ladder times the dimension’s weight. Computed from the scores, not written by the model; the model’s sentence under each is the action.
Use a concrete noun like a specific product name (e.g., ‘Fine-tuning API for small LLMs’) in the headline.
Replace abstract terms with concrete capabilities (e.g., ‘fine-tune and deploy small LLMs’).
Include a named user metric or audited success figure to corroborate claims.
All three together: 69 → 79. Every dimension one rung up: 79 (grade B). Ceiling is 95.
Built only from facts already on the page — not the company’s words. Shown so the gap between what was written and what could have been written is concrete rather than a number.
Five alternative headlines in five angles, a subhead, a call-to-action line, and a line-by-line edit for each of the five dimensions — written only from facts already on reopenly.com, gated like a rating (anything with a struck word is dropped, not sold). Delivered on a private page in about a minute.
Five dimensions on one ladder — 5 · 20 · 35 · 50 · 65 · 80 · 95 — weighted clarity 30%, specificity 20%, jargon 20%, proof 15%, call to action 15%. For each: the rung the page matched, the words that drove it, and what would reach the next rung. Full rubric →
The page uses an abstraction-heavy opening and a later sentence with a generic claim about an end-to-end platform, not a concrete product noun.
The end-to-end platform for small language models.
Next rungUse a concrete noun like a specific product name (e.g., ‘Fine-tuning API for small LLMs’) in the headline.
There are concrete numbers and pricing details tied to outcomes (cost, latency), which are testable.
Latency Milliseconds, not seconds A tuned small model decodes at about 25 ms per token on our production hardware
Contains abstract terms such as end-to-end, platform.
The end-to-end platform for small language models.
Next rungReplace abstract terms with concrete capabilities (e.g., ‘fine-tune and deploy small LLMs’).
There is one measurable data point about usage but no named customers or independent metrics.
Nothing on the page addresses this.
Next rungInclude a named user metric or audited success figure to corroborate claims.
A clear action exists ('Start free with 10 credits'), aligning with the promise of a product offer.
Start free with 10 credits
The end-to-end platform for small language models.Demonstrates reliance on abstraction rather than a concrete product noun.
Low-code fine-tuning for LLMsMentions a concrete capability (fine-tuning) but not a specific product category.
Small models, frontier results. · ReOpenly Skip to content Low-code fine-tuning for LLMs Small models, frontier results. The end-to-end platform for small language models. Tuned to your task, a small open model matches frontier accuracy at a fraction of the cost. Own your intelligence: private, compatible, no lock-in. Ever. [email protected] Start free with 10 credits No credit card · pay only for GPU minutes you use $0.03 / GPU·min flat Bring data your way: CSV import, API, or the built-in chat editor OpenAI-compatible API: swap one base URL and keep your code Weights are yours to download anytime. No lock-in. support-bot / Qwen3.5 9B dataset user Where is my order #4821? assistant Your order left our warehouse yesterday, tracking is in your account. + 1,238 rows · customer-support.csv Training · epoch 2/3 Train loss ↓ Where is my order? Your order shipped yesterday. Tracking appears under “Orders” within 24h of dispatch. Deployed Small models beat giants on your task · ~1/10 the cost · Faster responses · Your data trains only your models Your data CSV · API · UI → Datasets Upload CSV, push via API, or write examples in the editor Fine-tune One-click training on small open models Evaluate Live loss curves and checkpoints while it trains Serve Hosted inference in our cloud, or download the weights → Predictions REST API · OpenAI-compatible // do the math Estimate your savings Compare what you pay a big-model API today against a small model tuned to the same task, served on our
Strongest: Specificity 95. Weakest: Clarity 50. The 2 passes agreed within 3 points. The band 64–74 is where a re-score of this unchanged page would most likely land: on 42 dry re-scores (12 Oct 2026) the median move was 6 points and one in four moved more than 10. A grade only changes when a new score clears the boundary by 5.
No word in the headline itself was judged empty, yet jargon-free scores 65: the scorer found generic phrasing further down the page. Contains abstract terms such as end-to-end, platform.
Badges are issued to grades A and B. The share card for this page shows the headline with its jargon struck through; paste the URL anywhere that unfurls links.
Scored from the live homepage
on 2026-10-24 by gpt-5-nano,
2 independent passes, median taken.
Page snapshot 1d6afcced74e.
Rubric 7329c240 (current; what it says).
No person reviewed or edited this score. If the rubric and the score disagree, the rubric wins
and the page is re-scored — not hand-corrected.
Re-scored when the homepage text changes.