Text summarize

POST /api/text-summarize5 quota units per call · cache hits free

Summarize text into a shorter version using Llama 3.1 via Cloudflare Workers AI. Configurable length (1-5 sentences) and style (bullet points vs prose).

Summarize a long-form text into a shorter version using Llama-3.1-8B-Instruct via Cloudflare Workers AI. Useful for: - Email / meeting-notes triage. - Article / blog-post TL;DRs. - Pre-OCR summarisation of scanned documents (chain with document-ocr). `length` controls how many sentences (1-10). `style` picks between flowing prose (`prose`) and a bulleted list (`bullets`). The system prompt is implicit; override `prompt` for a fully custom instruction. **Cost: paid Cloudflare Workers AI inference.** Llama-3.1-8B is a mid-sized text model — each call charges against your account's Workers AI budget plus 5 wasm-tool quota units per call.

Inputs

Name	Type	Default	Description
text*	string	—	The text to summarise. Plain text, no markup.
length	number (1…10)	3	Target number of sentences (prose) or bullet points (bullets).
style	enum (prose \| bullets)	"prose"	Output style — running paragraph or a bulleted list.
prompt	string	—	Override the system prompt entirely. When set, `length` and `style` are ignored — the user prompt is just the source text.

Response

Modes: json, text. Cache: yes (24h TTL).

Every response carries x-cache (HIT / MISS / BYPASS), x-cache-signature (stable across identical inputs), and the rate-limit headers listed below.

Quota & limits

Cost per call	5 units against the 10,000-unit monthly quota. Cache hits are free — only cache misses decrement.
Burst limit	200 requests per rolling minute, shared across all tools.
Max request size	`128 KiB` (base64 inflates file payloads ~33% — the limit applies to the decoded bytes).
Max response size	`16 KiB`
Timeout	60s wall clock.
Rate-limit headers	`X-RateLimit-Limit`, `X-RateLimit-Remaining`, `X-RateLimit-Burst-Limit` on every response; `X-Quota-Warning` once 80% of the monthly quota is spent; `Retry-After` on 429s.

Errors

Errors are JSON with an error string; schema failures add a details object (Zod's flattened field errors).

Status	When	Body
400	Input fails the tool's schema (wrong type, missing required field, malformed base64) or the file bytes can't be parsed as the expected format.	{ "error": "invalid inputs", "details": { "fieldErrors": { … } } }
401	Missing or invalid X-Api-Key / X-Account-Id headers (or no session when called from the browser).	{ "error": "unauthorized" }
404	Unknown tool id, or calling /api/* on the website hostname instead of api.xtract.bot.	{ "error": "not found" }
413	Request body over the tool's max request size, or an image header declaring more megapixels than the tool's pixel budget.	{ "error": "…exceeds maxRequestBytes…" }
415	Content-Type isn't application/json on the json-body transport.	{ "error": "unsupported content-type; expected application/json" }
429	Monthly quota or the 200/min burst limit exhausted. Check Retry-After and the X-RateLimit-* headers.	{ "error": "monthly quota exceeded", "monthRemaining": 0, … }
5xx	Conversion engine failure. Safe to retry; if it persists, the input is hitting a bug — please report it.	{ "error": "…" }

Code samples

Built from the default example.


curl -X POST https://api.xtract.bot/api/text-summarize \
  -H "Content-Type: application/json" \
  -H "Accept: application/json" \
  -H "X-Account-Id: $XTRACT_ACCOUNT_ID" \
  -H "X-Api-Key: $XTRACT_API_KEY" \
  -d '{
  "text": "Cloudflare Workers AI lets developers run hosted-inference models from inside a Worker. Models include text generation (Llama, Qwen), vision-language (LLaVA), image generation (Flux, SDXL), and image classification (ResNet). Calls are metered per request — pricing varies by model size and the type of inference. Workers AI is one of three new Cloudflare AI products alongside Vectorize (vector storage) and AI Gateway (caching + observability for inference calls)."
}'