Models
A mixture of models, cross-checked.
Every frontier model (Claude, GPT, Gemini, Mistral, Grok) alongside our own, routed per sub-task and made to check each other. No single-model bias.
Explore Science vs Manusights
Manusights sells pre-submission reports per paper, from a free readiness scan to a human expert. Explore Science is the AI co-scientist: up to two hours of cross-checked analysis, a Calibre score whose weights are published, references checked against the claims they support, and every draft scored in a workspace your co-authors can join.
The short answer
Choose Explore Science when the paper is going to be revised: up to two hours of cross-checked analysis from a named mixture of frontier models, a scoring formula you can read down to the weights, and every draft scored side by side with a chat that has read the whole file. Choose Manusights for a cheap look at one manuscript, or when a deadline needs a human name on the feedback. Both go well past grammar into the science, and both now compare a revision against its prior round. If you are aiming at a selective journal over several drafts, Explore Science is built for that arc.
Compared July 2026 · from Manusights's own published materials
Choose Explore Science when
- Goes deeper: up to two hours of cross-checked analysis on a full review, against about thirty minutes for a Manusights full review and an hour or more for its $99 dossier.
- Names its engine: a disclosed mixture of frontier models (Claude, GPT, Gemini, Mistral, Grok) plus our in-house Explorer One, cross-checking each other; Manusights describes hundreds of frontier-LLM calls without naming a model.
- Publishes the arithmetic, not just the criteria: the scoring page gives the Calibre formula, its weights, and the per-field evidence hierarchies behind each score.
- Scores every version in a workspace your co-authors can join, showing which dimensions moved between drafts.
- Rosa, the co-scientist chat, is grounded in your paper, its review, and your wider body of work, so you can interrogate any finding; Manusights describes no chat interface.
Choose Manusights when
- Faster and lighter for a one-off: a free readiness scan in minutes and a full review in about thirty.
- Cheaper for a single paper: $39 against our $49 one-time review, with review packs down to about $28 each and a 60-day money-back guarantee.
- Human tiers Explore Science does not offer: a field-matched scientist from $1,000 in three to seven business days, or a Cell, Nature or Science editor at $1,500 to $1,800 in five to ten days with a call.
- Publishes a methods page with its dimensions, its 1 to 5 scales, and blind validation case studies against real reviewer reports.
- Ships the review as a Word document, plus free standalone tools that screen a bibliography for retractions and hijacked journals.
Side by side
| Capability |  | Manusights |
|---|---|---|
| What it reviews | Manuscripts in any field, plus protocol/proposal, editorial, and novelty reviews | Manuscripts: readiness scan, full review, and a Submission-Ready Dossier with a four-persona simulated review, AI-content scan, and journal strategy |
| Turnaround | Up to two hours on a full manuscript review; emailed when it lands | Readiness scan in minutes; full review in about thirty; dossier an hour or more |
| Review engine | A named mixture (Claude, GPT, Gemini, Mistral, Grok) plus our in-house Explorer One, cross-checking each other on the same manuscript | Hundreds of frontier-LLM calls in a multi-stage pipeline; no model named |
| Scoring | Calibre: a score out of 100 across eight categories, with sub-scores | Section scores of 1 to 5 with an overall recommendation and an A/B/C action list; its methods page lists six dimensions where its product pages name eight |
| Scoring method | Published down to the weights, plus the per-field evidence hierarchies behind each score | Publishes its dimensions, scales, and validation cases; not the weights that combine them |
| Reference checking | Every reference resolved against live databases and checked against the claim it is cited for; non-resolving entries flagged in your paper | Claim-level verification against CrossRef, PubMed, OpenAlex, Semantic Scholar, bioRxiv and medRxiv, plus retraction screening |
| Figure analysis | AI-generated-figure detection available as a plug-in | Vision-based parsing of every figure and image-embedded equation |
| Revisions | Every version scored, with the movement shown dimension by dimension in a shared workspace | Re-uploads reviewed against the prior round: what you addressed, what persists, what is new; no shared workspace described |
| Paper-grounded chat | Rosa answers follow-ups on this paper, its review, and your earlier work | No chat interface described; structured reports only |
| Human expert option | AI review only | Scientist from $1,000 in three to seven business days; Cell, Nature or Science editor at $1,500 to $1,800 in five to ten days |
| Training on your work | Does not train models on your manuscript; you can permanently delete it | States manuscripts are never used for training and are encrypted in transit and at rest; describes itself as an Anthropic Privacy Partner |
| Pricing | Free to start; $49 for one full review with no subscription, or monthly plans from $99 with a usage allowance | Free scan; $39 full review; $99 dossier; 3-pack $89, 5-pack $139; expert review $1,000 to $1,800; 60-day money back |
90% rate a review on par with or better than the human peer review they have had.
Two in three rate it better. From 304 head-to-head ratings by researchers who had received peer review on the same paper. These are our own users, self-reported, which is worth knowing when you weigh them.
Depth
Why a thorough review takes the time it takes.
Most tools optimize for speed. Explore Science runs the engine we built to conduct original research autonomously, and points it at your manuscript: it reads recent literature in your field, checks every reference, and works through the paper the way a careful reviewer does.
The time is the method, not an overhead on it.
How a paper is scoredTime spent thinking about your paper
How long a review actually takes
Single-pass LLM
~2 min
Other AI reviewers
5 to 20 min
Human peer review
(active reading time)
~30 min
Explore Science
up to 2 hrs
Industry observation; self-reported vendor figures; Publons peer-review time survey, 2018.
The engine
What actually reads your paper.
Three pieces of the architecture do the work behind every review.
Agents
Scientific agents that reason like the field.
Dozens of specialised agents work each review in parallel, each tuned to one dimension of analysis, spawning more when a section needs them.
Infrastructure
An in-house science stack.
Memory that holds your paper in context for the whole review, live verification of every citation, and novelty assessment. Built over years of autonomous science.
Where the extra time goes
Manusights shares a premise most tools do not: it goes past grammar into the science, running a multi-stage pipeline of frontier-LLM calls against a rubric its reviewer network co-designed, and it parses every figure and verifies citations along the way. Its ladder is built around price per paper: a free scan, a $39 full review in about thirty minutes, and a $99 dossier that runs an hour or more and simulates four reviewer personas.
Explore Science spends up to two hours on a single manuscript, letting a named mixture of frontier models and scientific agents check each other before anything reaches you. The same eight-category rubric applies whether the paper is molecular biology or condensed-matter physics, and the review arrives by email, so the extra time costs you nothing but patience.
Both publish something about the score
Manusights publishes a methods page: the dimensions it scores, the 1 to 5 scales, the pipeline stages, and case studies blind-tested against real reviewer reports. What it does not publish is how those dimensions combine into the number at the top.
Explore Science publishes the arithmetic as well as the criteria, including the weights and the per-field evidence hierarchies used to judge a study's design. An author deciding what to fix first can work backwards from the weighting to the fix that moves the score most.
What happens on the second draft
Manusights re-reads a revised upload against your prior round and tells you what you addressed, what still stands, and what is new; its 3-pack and 5-pack credits are priced for exactly that cycle. Each round arrives as its own Word document.
Explore Science keeps the paper in a workspace your co-authors can join, scores every version, and charts which dimensions moved between drafts, so a months-long revision has a visible arc rather than a stack of reports. Rosa, grounded in the paper and its review, lets you ask why a finding landed and what would move it.
Common questions
What does Manusights cost, and do I need a subscription with Explore Science?
Manusights charges nothing for a readiness scan, $39 for a full review, $99 for a Submission-Ready Dossier, and $1,000 to $1,800 for a human expert; packs of three and five bring the per-review price to roughly $28. Explore Science sells a single full-depth review for $49 with no subscription, or monthly plans from $99 with an allowance covering reviews and chat. For one paper Manusights is the cheaper ticket; across a revision cycle the allowance costs less per review than either.
I have reviewer comments back from a journal. Which one helps more?
Both re-read a revised draft in light of the previous round. Manusights returns each round as a separate Word document and sells credit packs for the cycle. Explore Science scores every version in one workspace, so you can see which dimensions moved and which objections are still live, and Rosa can be asked what would move a specific score next.
Does either tool work outside biomedicine?
Manusights calibrates its rubric against a network of reviewers drawn largely from oncology, cardiology, infectious disease, immunology, genomics, neuroscience, and metabolism, and its expert tiers are matched within those fields. Explore Science applies the same eight-category rubric to any field, from condensed-matter physics to ecology, with no field-matched human step to arrange.
Do Manusights or Explore Science train AI on my manuscript?
Neither says it does. Manusights states manuscripts are never used to train any model, encrypts them in transit and at rest, and describes itself as an Anthropic Privacy Partner. Explore Science does not train models on your manuscript and lets you permanently delete any paper from your workspace.
Claims about Manusights reflect its own published materials; everything about Explore Science reflects what the platform does today. Sources: Manusights, Manusights pricing, Manusights AI review, Manusights methods, Manusights expert services, Manusights FAQ. Something out of date? and we will fix it.
Already using Manusights? Put the same paper through a scored review.
Get started free