Stop hand-crafting prompts. Ship humanization.

Name: HumanizeAI API
Brand: HumanizeAI

The Humanize API. Multi-detector targeting, frozen keywords, style mirroring — all in one HTTP call. Same engine that powers humanizeai.chat.

Get an API key Read the docs

QuickstartPOST /v1/humanize

curl https://api.humanizeai.chat/v1/humanize \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Your AI-generated draft here.",
    "tone": "academic",
    "target_detector": "all",
    "frozen_keywords": ["mitochondria", "ATP synthesis"],
    "return_detection": true
  }'

Pick a tier

Pay-as-you-grow with no surprise charges. Soft caps available.

Free Dev

$0/forever

100 calls/mo · 5 RPM

Stealth model
All v1 endpoints
Community support

Get a key →

Builder

$29/month

5,000 calls/mo · 30 RPM

Pro humanizer modes
Frozen keywords + 3 styles
Webhook events
99.5% SLA
Email support

Start building →

Scale

$99/month

25,000 calls/mo · 120 RPM

Everything in Builder
Unlimited writing styles
Custom prompt modes
99.9% SLA
Priority support

Go to scale →

Enterprise

Customfrom $499/mo

Unlimited · 600+ RPM

Fine-tunes per customer
Dedicated SE
99.95% SLA + contract
Custom DPA + SOC 2
Self-serve key audit

Talk to sales →

Built for production

Things you actually need when shipping AI features — not just a /chat wrapper.

Multi-detector targeting

Pass target_detector: gptzero | turnitin | originality | copyleaks. Outputs are tuned to defeat the specific detector you care about.

Frozen keywords

Pin terms that must survive the rewrite verbatim. Critical for SEO + technical writing where exact phrasing matters.

Style mirroring

Pass a style sample or saved style_id. Output mirrors voice, sentence length, vocabulary level.

Detection breakdown included

Every response ships with a detection score and component breakdown (burstiness, vocabulary, perplexity) so you can decide whether to iterate.

Sane rate-limit headers

X-RateLimit-Limit / Remaining / Reset on every response. X-Usage-Period-Used / Limit / Reset for billing-period caps.

Webhooks

humanize.completed, usage.limit-warning, usage.limit-reached. Signed with your secret. Retried on 5xx.

Why developers ship on HumanizeAI

3.1s P50 latency on /humanize

99.9% uptime on Builder + Scale

1 HTTP call per humanization — no orchestration on your side

Stripe billing — cancel anytime, no contract until Enterprise

Ready to ship?

Start free. Upgrade when you need more calls. Read the docs to see every endpoint.

Get an API key