Add cited AI to your help center
Reduce repetitive tickets by embedding answers that link to source articles — so customers and agents can verify before they act.
Document Q&A · cited sources · tenant-isolated
Cited doc answers
Cited doc answers
in one API call
Upload PDFs and Markdown, test answers with page-level citations in the dashboard, then ship POST /v1/chat with one key. Runs on Oprag's AWS — isolated per company and project.
Same /v1/chat endpoint on Starter and Pro. Limits are queries, not API access.
POST /v1/chat X-Oprag-Key: sk_live_…
Your data stays yours. Documents and queries live on Oprag's AWS, isolated by company and project — encrypted at rest, never used to train shared models.
Stack Bedrock · S3 · Cognito · CloudFront · DynamoDB · Lambda
Security overview →How it works
From upload to production
three steps, one API key
Upload, verify citations in the test chat, then ship the same API to production.
01
Upload your knowledge
Drop in PDFs, Word docs, Markdown, or URLs. We parse, chunk, and index everything automatically — no pipeline to build.
Start for free →pricing.pdf
faq.md
guide.pdf
02
Test in the dashboard
Chat with your assistant before you ship. Every answer shows cited sources so you can verify accuracy instantly. No credit card on Starter.
Start free →What's the cancellation policy?
You can cancel any time from Account → Billing.
billing-faq.md · line 42
03
Ship via API
One API key, one endpoint. Embed the widget or call the REST API from your backend — same contract from sandbox to production.
View API docs →POST /v1/chat
X-Oprag-Key: sk_live_****
{
"question": "refund policy?"
} 200 OK with cited sources
Ready to test? Upload a PDF, ask one question, see the citation — free Starter includes the full API.
Start free — no credit cardGuides
How teams ship cited answers
Practical workflows from Oprag — REST integration, help-center embeds, and validation before launch.
Browse all guides →Validate document Q&A before you launch
A founder-friendly checklist for testing cited answers in the dashboard before you spend on traffic or engineering time.
Integrate document Q&A with a REST API
Upload docs, verify cited answers in the dashboard, then call one REST endpoint from your app — no RAG pipeline to build.
What you get
Four things that matter
when you ship doc Q&A
Depth over breadth. Each block is something you can verify in the dashboard or API docs.
API
Same endpoint from Starter to Enterprise
Same REST endpoint on every plan. Prototype on Starter, scale on Pro — one key, one schema.
POST /v1/chat
X-Oprag-Key: sk_live_****
{ "question": "..." }Citations
Every answer names its source
Document, page, passage — so users and auditors can verify. Not a bare LLM reply with a filename badge tacked on.
source: pricing-policy.pdf · p.4
confidence: highSecurity
Tenant-isolated by design
Documents and queries are scoped to your company and project on Oprag's AWS stack. Encrypted at rest in S3, retrieved via Bedrock — we don't use your content to train shared models.
tenant_id · project_id
S3 · Bedrock · encrypted at restIngestion
PDF, Markdown, DOCX, URLs
Chunking and re-sync handled automatically when sources change. Upload via dashboard or presigned S3 URLs from the API.
PDF · MD · DOCX · URL
auto-chunk · re-indexWhy teams switch
DIY RAG breaks in production
- Embedding pipelines that drift when docs update
- Answers with no source your team can verify
- API access gated behind $100+/mo before you can test integration
Built for
Teams who need answers
grounded in their own docs
Three common paths to the same outcome: verified answers from your own docs.
Support leads
Answer help-center questions with sources
Embed a cited widget in your help center. Every answer links to the source article or PDF page — so agents can verify before it reaches customers.
Embed widget · cited repliesPlatform engineers
One endpoint, no RAG pipeline
Index docs via dashboard or API. Same /v1/chat contract from sandbox to production — one key, one schema, no surprise paywalls.
1 API key · same schemaFounders & PMs
Validate before you scale spend
Test answer quality in the dashboard, then ship to users when citations look right. Full API access on Starter — no credit card required.
500 queries/mo on StarterPricing
Simple, predictable pricing
API on every plan
API access included from day one — even on Starter. Flat monthly tiers, no credit math.
Starter
Start for free Free / forever
Build and validate your chatbot — no credit card needed.
- 1 project
- Up to 50 documents
- Full REST API access incl.
- Cited answers
- Dashboard & testing
- 500 queries / month
- White-label widget
- Priority support
Most teams
Pro
Start Pro trial $49 / per month
Ship to real users with higher limits and branding control.
- 5 projects
- Unlimited documents
- Full REST API access incl.
- Cited answers
- Dashboard & analytics
- 20 000 queries / month
- White-label widget
- Priority support
Enterprise
Talk to us Custom / annual contract
For teams that need SLAs, SSO, and dedicated tenancy on Oprag's AWS.
- Unlimited projects
- Unlimited documents
- Full REST API access incl.
- Cited answers
- Custom analytics & BI
- Unlimited queries
- White-label widget
- Dedicated tenancy + SLA
Full API on every plan. No annual contract on Starter or Pro.
Frequently asked questions
Can I really use the API on the free plan?
Yes. Every plan — including Starter — includes full REST API access. Create a project, deploy once to get your API key, and call the same /v1/chat endpoint used in production. The free tier limits projects and monthly queries, not API availability.
What counts as a query?
One query is a single chat request to your project's knowledge base — whether from the dashboard test chat or your production API. Follow-up messages in the same conversation each count separately. Starter includes 500 queries per month; Pro includes 20,000.
How do I upgrade or cancel?
Upgrade to Pro anytime from your dashboard billing settings — changes take effect immediately. Downgrade or cancel with one click; you keep access through the end of your billing period. No contracts on Starter or Pro.
Where does my data live?
On Oprag's AWS infrastructure — not in your AWS account. Documents and queries are isolated by company and project. Encrypted at rest in S3; retrieval via Bedrock within our managed environment.
Can I see the API before signing up?
Yes. The full API reference is public at /docs — request shapes, auth headers, and response format. No login required to read it.
Is my data shared with any third-party AI?
Your documents are stored on Oprag's AWS infrastructure, isolated by company and project. Queries are processed through AWS Bedrock — we do not send your data to third-party model providers or use it to train shared models.
Recently shipped
2026.06.1 Isolated RAG API — Aurora, hybrid search & streaming chat
- Standalone RAG service at /v1/rag on api.dev.oprag.ai — separate Lambda from dashboard and embed paths, service-key authenticated
- POST /v1/rag/ingest and POST /v1/rag/ingest/batch — queue up to 50 documents per batch; GET /v1/rag/jobs/:jobId polls ingest status to indexed
- S3 ingest from tenant-scoped keys (documents/{tenantId}/{projectId}/…) with bucket validation and path-traversal rejection
- Local extraction for PDF, DOCX, and plain text before chunking — no client-side parsing required
Self-serve signup
Upload docs. Test citations. Ship one API key.
Starter includes the full REST API — 500 queries/month, no credit card.
No credit cardFull API on StarterTenant-isolated