Drop your contract here
or click to browse — PDF only, up to 25 MB
Selectable text required — password-protected and scanned PDFs not supported
What gets analyzed
AFileFix's Contract Review is a research aid. It reads your contract with AI, flags issues, and suggests negotiation language — but it doesn't replace a lawyer.
If your contract involves significant money, ongoing obligations, or potential disputes, have a lawyer review it before signing. Especially for: employment contracts you'll be bound to for years, vendor agreements with large dollar values, leases with multi-year terms, anything with personal liability or IP implications.
What Contract Review does well: it surfaces things you might not notice. It tells you what to ask about. It gives you starting language for negotiations. What it doesn't do: replace professional legal judgment for your specific situation, jurisdiction, and risk tolerance.
Upload a contract PDF. The AI reads the full document, identifies the contract type (vendor agreement, employment, NDA, lease, etc.), then returns a structured analysis:
Processing takes 15-30 seconds for typical contracts. The document is forwarded to Anthropic's Claude API for analysis and deleted from our server within minutes after processing. Anthropic retains API data for up to 30 days for safety review, then permanently deletes it; data is never used to train AI models.
The analysis is calibrated to common contract types:
Vendor and SaaS agreements — Limitation of liability scope and caps, IP ownership of work product, data ownership and portability, auto-renewal terms, termination rights, payment terms, warranties and disclaimers.
Employment contracts — Non-compete and non-solicit scope, severance triggers and amounts, IP assignment language, confidentiality term length, termination causes, dispute resolution forum.
NDAs — Term length, definition of confidential information (broad vs narrow), permitted uses, return/destruction obligations, jurisdiction.
Freelance and services agreements — Scope of work clarity, payment milestones and timing, IP ownership of deliverables, revision limits, kill fees, indemnification.
Leases — Rent escalation, renewal terms, security deposit terms, maintenance responsibilities, sublet restrictions, early termination penalties.
Generic commercial agreements — When the contract type isn't clearly identifiable, the AI applies a broader standard-clause checklist and flags significant deviations from common commercial terms.
Each flag has a severity rating reflecting how much attention it warrants:
High — A clause that significantly disadvantages you or creates substantial risk. Common examples: unlimited liability, perpetual IP assignment, automatic renewal without notice opportunity, broad non-compete without geographic limit. Don't sign without addressing these.
Medium — A clause that's tilted against you but is often negotiable. Common examples: short notice periods for termination, narrow warranty language, restrictive payment terms. Worth pushing back on.
Low — A clause that's worth knowing about but probably standard. Common examples: governing law selection, force majeure scope, dispute resolution forum. Note for awareness; usually accept.
A few things to know about the severity ratings: they reflect the AI's reading of the clause language and contract type. They don't account for your specific leverage in the negotiation, your industry norms, or your appetite for risk. A “High” flag in a vendor contract from a Fortune 500 may be impossible to negotiate; the same flag in a contract from a small vendor is probably negotiable.
When Contract Review identifies a significant issue, it includes proposed negotiation language — text you can copy directly into a negotiation email or pass to your legal team.
These suggestions are starting points, not finished work product. They're written in standard contract language and reflect common approaches to common problems. They're not jurisdiction-specific (US contracts and UK contracts handle some issues differently), they're not pre-vetted by a lawyer, and they don't account for your specific business relationship with the counterparty.
How to use redlines well: paste the suggested language into your negotiation, then modify it for your tone, your relationship, and your priorities. If your lawyer is involved, share the AI's suggestions as research context — your lawyer will refine them based on actual legal expertise.
How NOT to use redlines: copy verbatim into a signed contract without legal review. The suggestions are research output, not finished legal work.
Best results: Standard text-based PDFs of common contract types (vendor, employment, NDA, lease, services). Modern contracts created in Word and exported to PDF — these have clean text the AI can read accurately.
Modest results: Heavily annotated contracts with markup, redlines from previous rounds of negotiation, or unusual formatting. The AI can still extract meaningful analysis but accuracy on specific clauses drops.
Not supported well:
Yes, for most commercial agreements. When the AI can't identify a specific contract type from the supported list, it falls back to a general commercial checklist and flags significant deviations from common terms. The analysis is less calibrated than for known types, but you'll still get risk flags, missing-clause notes, and key dates. For unusual contract types, expect more general output and treat the analysis as a starting point rather than a thorough review.
Yes. Run your own draft through Contract Review before sending it to the counterparty — the AI surfaces things you might have missed, weak clauses you could strengthen, or terms you forgot to include. Think of it as a research-grade self-check, not a substitute for legal drafting.
The AI checks against calibrated clause sets per contract type. If your concern is a non-standard clause not in those sets, it may be analyzed in the general output but not specifically flagged. For specific clauses you want to investigate, use Chat with PDF — ask Claude direct questions about what your contract says on that specific topic.
For ad-hoc questions about specific clauses, try Chat with PDF. For AI-processing details and Anthropic's API retention terms, see our Privacy page.
Free, no signup. Get risk flags, clause checklist, key dates, and negotiation language.