Multi-Agent AI Review

Mentor-Level AI
Guards Your
Thesis

Upload a Word (.docx) file, confirm scope in chat, then let specialized agents work in parallel. Pricing follows your selections and word-count tier (shown in-app); structured reports usually arrive within minutes. Reviews surface actionable notes across layout semantics, argument flow, reference style, and optional AI-trace hints so you can revise before submission—the output is advisory guidance, not a substitute for institutional or journal peer review.

< 5min
Typical runtime
4
Co-op agents
120k
Chinese chars max
✓ No card required✓ Edu users can claim trial credits✓ Credits never expire
ThesisAI · Review in progressLive
📄
Research…
Master thesis · Computer vision · Word
Running
🤖 ThesisAI
Detected a computer-vision master thesis. Focus on logic, format, or run all four agents?
Logic firstInclude formatAll four agents
Prioritize the logic pass
Agent progress
📋
Format100%
🧠
Logic68%
🔎
AI trace42%
📚
References31%
Core Features

Not just checking,deep understanding

Built for Chinese thesis workflows: pick modules in chat, read results in tabs—no vague “does everything” claims.

🕵️

Multi-agent collaboration

Dedicated agents for format, logic, AI trace, and references work in parallel without blocking each other. Adjust what runs before you start.

FormatLogicAI traceReferences
💬

Conversational setup

After upload, confirm degree type, priorities, and format notes; the dashboard shows an estimated credit cost before you start.

Degree typeFormat notesCost estimate
📊

Visible progress

A live agent timeline shows queueing, running, and completion; streaming updates reduce the “black box” wait.

Agent statusStreamingTraceable
📋

Format & layout

Uses your format brief plus Word layout semantics (headers/footers, page numbers, etc.). Reference styling is checked against common GB/T 7714-2015 issues.

Format briefLayout semanticsGB/T 7714
🧠

Logic & argument

Paragraph-level clues and argument-chain checks—abstract/conclusion alignment, leaps in reasoning, common defense risks.

Paragraph cluesArgument chainDefense risks
🔎

AI trace hints

Highlights AI-like phrasing and expression risks across the body text so you can rewrite before submission (advisory only, not an academic verdict).

Trace hintsStyle risksParagraph-level
Workflow

Four steps tomentor-level pre-review

From upload to report, fully automated—no tedious forms

📤1

Upload thesis

Drop a Word (.docx) file; we parse structure and word count. Optional research field. PDF and legacy .doc are not supported.

Word · Max 50MB
💬2

Conversational config

Confirm degree level, focus, and your format notes. You’ll see the cost estimate in the dashboard before the run starts.

Bachelor / Master / PhD
🚀3

Multi-agent collaboration

Selected agents run in parallel with live status for each one.

Usually minutes
📊4

Review & export

Results are split across Format, Logic, AI trace, and References tabs. Export a Markdown bundle from the report view, or print to PDF from the browser.

Markdown export
Review report · Deep learning image recognition
12
Sample finding count (not a score)
Format3items
Logic4items
References2items
AI trace3items
Core issues · Priority suggestions
HighSection 3.3 reads circular—reorder the analysis chain and add an explicit contrast or control.Sec. 3.3
MediumSeveral paragraphs share highly templated phrasing; rewrite with concrete numbers and qualifiers to reduce false AI positives.Body
MediumReference [12] may be missing year/volume fields—double-check against GB/T 7714-2015.Refs
LowFigure 2-3 is cited as “Fig. 2-4” in the text—align numbering with the first in-text mention.Sec. 2.2
📄
4
Co-op agents
50MB
Max file size
< 5min
Typical runtime
🏆
GB/T
Ref. style cues

Students love ThesisAI

Early users & beta notes (lightly anonymized)

The references tab flagged missing fields faster than eyeballing a PDF export. AI-trace hints also pushed me to rewrite a few templated paragraphs.

🎓
Wang
Huazhong Univ. · Master

Logic review caught a reasoning leap I’d glossed over before my defense. Progress per dimension was easy to follow in the dashboard.

🔬
Chen
Peking Univ. · PhD

Uploading docx and toggling modules in chat feels natural. Exporting Markdown back into Word for formatting tweaks is quick.

📚
Li
Zhejiang Univ. · Bachelor
Credit Plans

Transparent pricing,pay as you go

Prepaid credits, no subscription. Balance never expires; each run bills only the modules you select, with word-count tiers—see the estimate in-app before you start.

单次体验
Single Pass
¥9.90¥29.90/ Starter

Try modular review and the credit model

Credits added to your wallet, deducted per run
Pick any combination of four modules
Markdown export from the report view
Word (.docx) up to 50MB per file
Support tickets when enabled by ops
专业套餐
Pro Bundle
💎
¥800¥1495/ Team
Save 40%

For labs and teaching assistants helping many students

High-volume credit pack
Automatic partial refunds for failed modules
Exports and history retained
Same per-thesis word cap as other tiers
Admin/ticket workflows when enabled
Billing notes:
What you pay forOnly enabled modules
Word tiersAdjust per-module prices (see top-up page)
FailuresFailed modules refund automatically
Theses over the configured word cap cannot start; check the dashboard for the live limit
FAQ

You might ask

Questions? This section covers supported thesis types and upload limits, credit expiry, privacy and data handling, failure/refund behavior, exporting reports, and payment methods—quick checks before you sign up or start a run. If something is not covered, use the in-dashboard support path to reach operations.

🚀

Before submission,
let AI check your thesis first

Edu users can claim credits from time to time. After you upload Word, review the cost estimate for this review, then start when you’re ready. No card required.

✓ No monthly fee✓ Credits never expire✓ WeChat/Alipay