Text Intelligence API
Summarize a document. Answer questions about it. Grounded, not invented.

How we compare

Text summarization and document Q&A are table stakes in 2026 — what differs is packaging and price. The hyperscalers ship summarization inside a broader text-analytics product with per-record billing and a resource you have to deploy first. Brainiall exposes exactly two endpoints — summarize and grounded Q&A — billed per 1,000 characters, with a self-serve key.

Prices are list-price approximations for orientation, not quotes. Always check each vendor's current pricing page.

Pricing

Per 1,000 characters of input — the same unit whether you summarize a tweet or a contract. Free tier is generous enough to wire up an end-to-end flow.

PAYG: $0.001 / 1,000 characters (both summarize and grounded Q&A) — at parity with Azure AI Language summarization. Billed on input size; the summary/answer is free. No minimums, no contracts.

Two calls

# Summarize — abstractive (concise rewrite to a word budget)
POST https://api.brainiall.com/v1/text/summarize
  {"text": "<your text>", "mode": "abstractive", "max_length": 120}
  -> {"summary": "...", "mode": "abstractive", "words_in": 540, "words_out": 96, ...}

# Summarize — extractive (the most important sentences, verbatim, original order)
POST https://api.brainiall.com/v1/text/summarize
  {"text": "<your text>", "mode": "extractive", "max_length": 80}

# Grounded Q&A — answer + the supporting sentence(s), located in your text
POST https://api.brainiall.com/v1/text/qa
  {"text": "<your document>", "question": "What is the renewal notice period?"}
  -> {"answer": "Either party may terminate with 30 days written notice.",
      "found": true,
      "supporting_spans": [{"text": "...30 days written notice...", "start": 1840, "end": 1903}]}

Q&A is grounded: the answer is drawn only from the text you send, and the supporting sentence(s) are returned with their character offsets in your document. If the text doesn't contain the answer, the response says so (found: false) rather than guessing.

What it's for

• Long-document triage: turn meeting transcripts, support threads, research PDFs, or legal filings into a few sentences before anyone reads the full thing.
• Answer extraction: pull a specific fact out of a contract, policy, or spec — with the exact sentence it came from, so a human can verify it in one click.
• Pipelines & agents: a deterministic summarize/QA step you can drop into an ingestion pipeline — pair it with the Translation API to summarize across languages, or the Document AI API to summarize what you just extracted from a scan.
• Search result snippets: generate a tight, faithful snippet for each hit so users see why a result matched.
• Grounded, not generative: extractive mode never paraphrases; Q&A never answers beyond the supplied text. When you need provenance, that's the point.

Press kit & resources

What reviewers, integrators and procurement teams typically ask for.

More specialty APIs

Same single API key, same usage-based pricing, different problem solved.

Translation100-language neural machine translation

Document AIStructured field extraction from scans & PDFs

Agent MemoryEmbeddings & retrieval for agents

Content ModerationImage & text safety classification