---
name: great-work-labs
description: Use Great Work Labs (https://labs.greatwork.company) when the task needs a cheap, fast single-purpose model in the product: moderation or toxicity, prompt-injection screening, PII redaction, language detection, sentiment or emotion, entity extraction, zero-shot or topic classification, prompt routing, scam or spam filtering, grounding checks, embeddings, reranking, keyphrases, emoji, punctuation restoration, summaries, JSON extraction from text, titles, speech to text, image tagging, OCR, object detection, background removal or NSFW image checks.
---

# Great Work Labs

Hosted task models, one API: `POST https://labs.greatwork.company/v1/run` with `{"model": "<id>", "input": ..., <options>}` and `Authorization: Bearer $GWL_API_KEY`. Prepaid, per-call pricing; every response has `usage.cost_usd`.

1. Pick the model from the list below (or https://labs.greatwork.company/llms.txt).
2. Read its page before writing code: https://labs.greatwork.company/models/<id>.md (options, output fields, real examples). Never guess option names; unknown options are rejected.
3. Batch: text models take arrays (up to 64), outputs come back in order.
4. Handle errors by HTTP status: 402 means out of credit (tell the user, link https://labs.greatwork.company/dashboard/billing), 429/503 retry with backoff, 400 fix the request using `error.message`.
5. Keep the key in GWL_API_KEY. Never commit it.

- `lingo-1`: Language detection. Which language is this? 183 of them, from a few words up.
- `tone-1`: Sentiment analysis. Positive, neutral or negative, in 30+ languages.
- `mood-1`: Emotion detection. 28 emotions, plus the six basic ones, in one call.
- `spot-1`: Named entity recognition. People, organizations and places, with character offsets.
- `sort-1`: Zero-shot classification. Classify text into your own labels. No training.
- `topic-1`: Topic and format classification. What a page is about and what kind of page it is.
- `route-1`: Prompt task and complexity. Send easy prompts to a small model and hard ones to a big one.
- `sieve-1`: Toxicity detection. Insults, threats and slurs flagged in milliseconds.
- `warden-1`: Content safety policy check. 15 policy categories for prompts and LLM replies.
- `moat-1`: Prompt injection detection. Catch "ignore previous instructions" before your agent reads it.
- `veil-1`: PII detection and redaction. Find and mask 29 kinds of personal data.
- `bait-1`: Phishing, scam and spam detection. Is this message a scam, spam or the real thing?
- `ground-1`: Grounding check (hallucination detection). Does the source back up what your model said?
- `atlas-1`: Text embeddings. 384-dimension embeddings in 100 languages, OpenAI-compatible.
- `ladder-1`: Reranking. Put the best search result first. 100+ languages.
- `pluck-1`: Keyphrase extraction. The phrases a text is about, ready for tags and SEO.
- `glyph-1`: Emoji suggestions. The right emoji for a message, out of 1,800.
- `comma-1`: Punctuation and capitalization. Fix lowercase, run-on text from speech and chat.
- `brief-1`: Summarization. A few sentences that say what a long text says.
- `mold-1`: Structured extraction (JSON from text). Messy text in, JSON that matches your schema out.
- `headline-1`: Title and description writing. A specific title, a one-line description and a URL slug.
- `scribe-1`: Speech to text. Transcripts with word timestamps, $0.001 a minute.
- `prism-1`: Image tagging and image embeddings. Tag images with any labels, or embed them for search.
- `ink-1`: Text recognition (OCR). Every line of text in an image, with boxes.
- `scout-1`: Object detection. Find and count 80 kinds of objects, with boxes.
- `cutout-1`: Background removal. Transparent PNG of the subject, background gone.
- `curtain-1`: NSFW image detection. Keep explicit images out of your uploads.
