Your documents,
answered in milliseconds.

Teams using ContextIQ

• Arcadia Health
• Stormfront Analytics
• Brio Legal
• Novela Publishing
• Tangram AI
• Cascade Finance

Performance benchmarks

• < 200ms p99 response time
• 99.97% uptime (12-mo avg)
• 4 model backends

• Multi-model switching

  Choose GPT-4o, Claude Sonnet, Gemini Pro, or Mistral Large per request. No lock-in, no extra config.

• Citation-aware answers

  Every response traces to its source — page, paragraph, exact passage. Not hallucinated references.

• Sub-200ms p99

  Streaming plus caching mean your users never wait. Benchmark table available in the docs.

• Eval suite built in

  Score every response with faithfulness, relevance, and groundedness metrics. Catch regressions before prod.

• Structured output

  Get JSON, Markdown, or plain text. Schema-validated responses. Works with any downstream pipeline.

• Webhooks + SDK

  Node.js, Python, Go SDKs. Webhook on async jobs. OpenAPI spec for self-generated clients.

1. 01

   Upload or link your document

   PDF, URL, Markdown, or plain text. Up to 500 pages per request. We handle parsing, chunking, and embedding.

   curl -X POST https://api.contextiq.dev/v1/ingest \ -H "Authorization: Bearer $CIQ_KEY" \ -F "file=@q4-roadmap.pdf" # → { "doc_id": "doc_9f3k2", "pages": 24 }
2. 02

   Choose your model and output format

   Switch models per request. Specify a JSON schema or get Markdown by default. Every call is stateless.

   curl -X POST https://api.contextiq.dev/v1/query \ -H "Authorization: Bearer $CIQ_KEY" \ -d '{ "doc_id": "doc_9f3k2", "question": "What are the top 3 priorities in Q4?", "model": "claude-sonnet-4", "format": "json" }'
3. 03

   Receive answers with citations

   The endpoint returns a structured object: summary, Q&A pairs, citations array, latency stats. Ready to pipe into your product.

   { "answer": "The top 3 priorities are: ...", "citations": [ { "page": 4, "text": "Priority 1: Ship eval pipeline..." }, { "page": 11, "text": "Priority 2: Enterprise SSO launch..." }, { "page": 17, "text": "Priority 3: Latency SLA reduction..." } ], "model": "claude-sonnet-4", "latency_ms": 187 }