Documentation | CHEVORA API Reference

Everything you need to integrate with CHEVORA AI API — from keys to production-ready scenarios.

Quick Start

Get a key, send a request, and deploy to production in 3 minutes.

1. Create a Project

2. Make a Request

curl https://api.chevora.ai/v1/text/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "AI infrastructure positioning",
    "max_tokens": 180,
    "temperature": 0.6
  }'

3. Handle Response

{
  "id": "gen_abc123",
  "object": "text_completion",
  "created": 1707264000,
  "model": "llama-3-70b",
  "choices": [{
    "text": "CHEVORA — European AI platform...",
    "finish_reason": "stop",
    "index": 0
  }],
  "usage": {
    "prompt_tokens": 10,
    "completion_tokens": 45,
    "total_tokens": 55
  }
}

Authentication

Use Bearer tokens for all requests. Keys are created in the dashboard → API Keys.

chevora_sk_1234567890abcdef

Authorization: Bearer chevora_sk_YOUR_KEY_HERE

Never pass keys in frontend code. Use them only on the server.

Text Generation API

POSThttps://api.chevora.ai/v1/text/generate

Private text generation with enterprise-grade isolation. EUR 0.75 per 1M tokens.

Parameters

prompt (string, required) — Your input text
max_tokens (integer, default: 150) — Maximum output length
temperature (float, default: 0.6) — Creativity level (0.0-1.0)
model (string, optional) — llama-3-70b, gpt-4, claude-3.5

Example: cURL

curl -X POST https://api.chevora.ai/v1/text/generate \
  -H "Authorization: Bearer chevora_sk_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Write a product description for AI platform",
    "max_tokens": 200,
    "temperature": 0.7
  }'

Example: Python

import requests

response = requests.post(
    "https://api.chevora.ai/v1/text/generate",
    headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
    json={
        "prompt": "Write a product description for AI platform",
        "max_tokens": 200,
        "temperature": 0.7
    }
)

result = response.json()
print(result["choices"][0]["text"])

Example: Node.js

const response = await fetch('https://api.chevora.ai/v1/text/generate', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer chevora_sk_YOUR_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    prompt: 'Write a product description for AI platform',
    max_tokens: 200,
    temperature: 0.7
  })
});

const data = await response.json();
console.log(data.choices[0].text);

Response (200 OK)

{
  "id": "gen_4f8b2c1a9e",
  "object": "text_completion",
  "created": 1707523200,
  "model": "llama-3-70b",
  "choices": [
    {
      "text": "CHEVORA AI Platform delivers enterprise-grade AI infrastructure with EU-only data residency. Built on AMD MI300X clusters in Slovakia, our platform combines sovereign computing with industrial-scale performance. Run LLM inference, training, and fine-tuning without data leaving European jurisdiction.",
      "index": 0,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 58,
    "total_tokens": 70
  }
}

Try it now?

Open playground ->

Translation API

POSThttps://api.chevora.ai/v1/translate

Neural translation with industry glossaries. 95+ languages. EUR 0.75 per 1M tokens.

Parameters

text (string, required) — Text to translate
source_lang (string) — Auto-detect if omitted
target_lang (string, required) — Target language code
glossary_id (string, optional) — Custom terminology

Example: Python

import requests

response = requests.post(
    "https://api.chevora.ai/v1/translate",
    headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
    json={
        "text": "AI infrastructure for regulated industries",
        "source_lang": "en",
        "target_lang": "de",
        "glossary_id": "gloss_banking_terms"
    }
)

translation = response.json()
print(translation["translated_text"])

Response (200 OK)

{
  "id": "trans_k9m2p4",
  "object": "translation",
  "created": 1707524000,
  "source_lang": "en",
  "target_lang": "de",
  "translated_text": "KI-Infrastruktur für regulierte Branchen",
  "detected_lang": "en",
  "confidence": 0.98
}

Document Context Reduction API

POSThttps://api.chevora.ai/v1/reduce-context

Intelligent segmentation of long documents. Compress 100K+ tokens. EUR 0.75 per 1M tokens.

Parameters

document (string, required) — Long text to reduce
target_length (integer) — Max output tokens
preserve_structure (boolean) — Keep headings/sections

Example: Node.js

const response = await fetch('https://api.chevora.ai/v1/reduce-context', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer chevora_sk_YOUR_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    document: longText,
    target_length: 2000,
    preserve_structure: true
  })
});

const data = await response.json();
console.log(data.reduced_text);

Response (200 OK)

{
  "id": "ctx_r3d8c2",
  "object": "context_reduction",
  "created": 1707524200,
  "original_tokens": 45000,
  "reduced_tokens": 1950,
  "compression_ratio": 0.043,
  "reduced_text": "Summary of key sections..."
}

Invoice OCR API

POSThttps://api.chevora.ai/v1/ocr/extract

Structured invoice extraction. 99.2% accuracy. EUR 1.50 per 1K pages.

Parameters

file (binary, required) — PDF, PNG, or JPG file
document_type (string) — invoice, receipt, contract
output_format (string) — json (default), csv

Example: Python

import requests

with open('invoice.pdf', 'rb') as f:
    response = requests.post(
        "https://api.chevora.ai/v1/ocr/extract",
        headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
        files={"file": f},
        data={"document_type": "invoice"}
    )

data = response.json()
print(f"Invoice #: {data['invoice_number']}")
print(f"Total: {data['total_amount']} {data['currency']}")

Response (200 OK)

{
  "id": "ocr_9k3m2n1p",
  "object": "invoice_extraction",
  "created": 1707523400,
  "invoice_number": "INV-2026-0042",
  "invoice_date": "2026-02-08",
  "due_date": "2026-03-08",
  "vendor": {
    "name": "ACME Corporation",
    "address": "123 Business St, Vienna, Austria",
    "vat_number": "ATU12345678"
  },
  "line_items": [
    {
      "description": "AI Platform Subscription",
      "quantity": 1,
      "unit_price": 550.00,
      "total": 550.00
    }
  ],
  "subtotal": 550.00,
  "tax_amount": 110.00,
  "total_amount": 660.00,
  "currency": "EUR",
  "confidence_score": 0.992
}

Classification API

POSThttps://api.chevora.ai/v1/classify

Real-time content classification. 99.7% accuracy. EUR 0.75 per 1M tokens.

Parameters

text (string, required) — Text to classify
taxonomy (array, required) — Custom labels
multi_label (boolean) — Allow multiple labels

Example: Node.js

const response = await fetch('https://api.chevora.ai/v1/classify', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer chevora_sk_YOUR_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    text: 'Need help with invoice payment issue',
    taxonomy: ['billing', 'support', 'sales', 'technical'],
    multi_label: false
  })
});

const data = await response.json();
console.log(data.labels);

Response (200 OK)

{
  "id": "clf_x7y8z9",
  "object": "classification",
  "created": 1707523600,
  "labels": [
    {
      "label": "billing",
      "confidence": 0.94,
      "rank": 1
    },
    {
      "label": "support",
      "confidence": 0.78,
      "rank": 2
    }
  ],
  "primary_label": "billing"
}

Receipt OCR API

POSThttps://api.chevora.ai/v1/ocr/receipt

Extract data from retail receipts. Loyalty programs, fraud detection. EUR 1.50 per 1K pages.

Parameters

file (binary, required) — Receipt image
extract_items (boolean) — Parse line items
detect_fraud (boolean) — Anomaly detection

Example: cURL

curl -X POST https://api.chevora.ai/v1/ocr/receipt \
  -H "Authorization: Bearer chevora_sk_YOUR_KEY" \
  -F "file=@receipt.jpg" \
  -F "extract_items=true"

Response (200 OK)

{
  "id": "rcpt_x4y2z8",
  "merchant_name": "SuperMarket",
  "date": "2026-02-10",
  "total": 45.80,
  "currency": "EUR",
  "items": [
    {"name": "Milk 1L", "price": 1.20, "qty": 2},
    {"name": "Bread", "price": 2.50, "qty": 1}
  ],
  "payment_method": "card",
  "fraud_score": 0.02
}

Legal Document OCR API

POSThttps://api.chevora.ai/v1/ocr/legal

Extract clauses with structure awareness. Signature detection. EUR 1.50 per 1K pages.

Parameters

file (binary, required) — Legal document PDF
extract_clauses (boolean) — Parse contract clauses
detect_signatures (boolean) — Find signatures

Example: Python

with open('contract.pdf', 'rb') as f:
    response = requests.post(
        "https://api.chevora.ai/v1/ocr/legal",
        headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
        files={"file": f},
        data={
            "extract_clauses": True,
            "detect_signatures": True
        }
    )

data = response.json()
print(f"Clauses: {len(data['clauses'])}")
print(f"Signatures: {data['signature_count']}")

Response (200 OK)

{
  "id": "legal_p9k3m2",
  "document_type": "service_agreement",
  "parties": ["Company A GmbH", "Company B AG"],
  "clauses": [
    {
      "type": "payment_terms",
      "text": "Payment within 30 days...",
      "page": 3
    }
  ],
  "signature_count": 2,
  "effective_date": "2026-01-15"
}

Logistics Document OCR API

POSThttps://api.chevora.ai/v1/ocr/logistics

CMR/BoL parsing. Customs compliance automation. EUR 1.50 per 1K pages.

Parameters

file (binary, required) — Bill of lading / CMR
document_type (string) — cmr, bol, customs
validate_hs_codes (boolean) — Check HS codes

Example: Node.js

const formData = new FormData();
formData.append('file', fs.createReadStream('cmr.pdf'));
formData.append('document_type', 'cmr');

const response = await fetch('https://api.chevora.ai/v1/ocr/logistics', {
  method: 'POST',
  headers: {'Authorization': 'Bearer chevora_sk_YOUR_KEY'},
  body: formData
});

const data = await response.json();
console.log(data.cargo_description);

Response (200 OK)

{
  "id": "log_m2n8p4",
  "document_type": "cmr",
  "consignor": "Sender Company",
  "consignee": "Receiver GmbH",
  "cargo_description": "Electronics parts",
  "weight_kg": 450,
  "hs_code": "8517.62",
  "origin": "DE",
  "destination": "AT"
}

Legal RAG Query API

POSThttps://api.chevora.ai/v1/legal-rag/query

Query legal knowledge base with citations. Contract search, precedent analysis. EUR 0.75 per 1M tokens.

Parameters

query (string, required) — Legal question
collection (string, required) — Knowledge base ID
top_k (integer, default: 5) — Results count
include_citations (boolean) — Return sources

Example: Python

response = requests.post(
    "https://api.chevora.ai/v1/legal-rag/query",
    headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
    json={
        "query": "What are GDPR data retention requirements?",
        "collection": "eu_regulations",
        "top_k": 3,
        "include_citations": True
    }
)

result = response.json()
print(result["answer"])
for citation in result["citations"]:
    print(f"Source: {citation['document']}")

Response (200 OK)

{
  "id": "rag_k3m9p2",
  "object": "legal_rag_query",
  "created": 1707524600,
  "answer": "GDPR requires data retention only as long as necessary for processing purposes...",
  "citations": [
    {
      "document": "GDPR Article 5(1)(e)",
      "page": 12,
      "relevance": 0.94
    }
  ],
  "confidence": 0.91
}

HR Policy Search API

POSThttps://api.chevora.ai/v1/hr-search

Search internal HR policies. Role-based access, change tracking. EUR 0.75 per 1M tokens.

Example: Node.js

const response = await fetch('https://api.chevora.ai/v1/hr-search', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer chevora_sk_YOUR_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    query: 'vacation policy for employees',
    department: 'engineering'
  })
});

const data = await response.json();
console.log(data.answer);

Enterprise Knowledge Search API

POSThttps://api.chevora.ai/v1/enterprise-search

Cross-platform knowledge search. Integrations with Confluence, SharePoint, and Notion. EUR 0.75 per 1M tokens.

Example: cURL

curl -X POST https://api.chevora.ai/v1/enterprise-search \
  -H "Authorization: Bearer chevora_sk_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "API authentication best practices",
    "sources": ["confluence", "notion"],
    "top_k": 5
  }'

Compliance Analytics API

POSThttps://api.chevora.ai/v1/compliance/check

Monitor regulatory requirements. Checks for MiCA, GDPR, and AI Act. EUR 0.75 per 1M tokens.

Example: Python

response = requests.post(
    "https://api.chevora.ai/v1/compliance/check",
    headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
    json={
        "document": contract_text,
        "regulations": ["gdpr", "ai_act"],
        "industry": "financial_services"
    }
)

compliance = response.json()
print(f"Risk score: {compliance['risk_score']}")

Retail Loss Prevention API

POSThttps://api.chevora.ai/v1/vision/loss-prevention

Real-time theft detection. Privacy-first video analytics. Price: EUR 550 per store per month.

Parameters

video (file or stream, required) — Store camera video
sensitivity (string, optional) — "low" | "medium" | "high" (default: "medium")
alert_webhook (string, optional) — URL for real-time theft alerts
zones (array, optional) — High-risk zones to monitor
privacy_mode (boolean, optional) — Face anonymization (GDPR) (default: true)

Example: Python

with open('store_video.mp4', 'rb') as video:
    response = requests.post(
        "https://api.chevora.ai/v1/vision/loss-prevention",
        headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
        files={"video": video},
        data={
            "sensitivity": "high",
            "alert_webhook": "https://your-store.com/loss-alerts",
            "privacy_mode": True
        }
    )

alerts = response.json()
for alert in alerts["detections"]:
    print(f"Time: {alert['timestamp']}, Risk: {alert['risk_level']}")
    print(f"Location: {alert['store_zone']}, Confidence: {alert['confidence']}")

Response (200 OK)

{
  "analysis_id": "loss_7x9k2m8p",
  "timestamp": "2026-02-11T16:24:18Z",
  "duration_minutes": 10,
  "detections": [
    {
      "timestamp": "2026-02-11T16:28:34Z",
      "risk_level": "high",
      "confidence": 0.89,
      "store_zone": "Electronics Section",
      "behavior_type": "concealment_detected",
      "video_clip_url": "https://storage.chevora.ai/loss/clip_abc123.mp4",
      "alert_sent": true
    },
    {
      "timestamp": "2026-02-11T16:31:12Z",
      "risk_level": "medium",
      "confidence": 0.72,
      "store_zone": "Cosmetics Aisle",
      "behavior_type": "suspicious_loitering",
      "alert_sent": false
    }
  ],
  "total_incidents": 2,
  "false_positive_rate": 0.08,
  "privacy_compliance": "GDPR - all faces anonymized",
  "usage": {
    "video_minutes_processed": 10,
    "cost": "€0.50"
  }
}

💰 ROI: The average retail store prevents EUR 24K-38K in annual losses (market shrinkage 1.5-2%). API cost: EUR 550/month = 96% ROI.

Footfall Analytics API

POSThttps://api.chevora.ai/v1/vision/footfall

Track customer flow. Heatmaps and dwell-time analytics. Price: EUR 550 per store per month.

Parameters

video (file or stream, required) — Entrance/sales floor camera video
zones (array, optional) — Zones to track (default: auto-detect)
generate_heatmap (boolean, optional) — Generate heatmap (default: true)
track_demographics (boolean, optional) — Age/gender estimation (optional, GDPR-compliant)
interval (string, optional) — "realtime" | "hourly" | "daily" (default: "hourly")

Example: Node.js

const formData = new FormData();
formData.append('video', fs.createReadStream('entrance.mp4'));
formData.append('generate_heatmap', 'true');
formData.append('track_demographics', 'false');

const response = await fetch('https://api.chevora.ai/v1/vision/footfall', {
  method: 'POST',
  headers: {'Authorization': 'Bearer chevora_sk_YOUR_KEY'},
  body: formData
});

const data = await response.json();
console.log(`Traffic count: ${data.total_visitors}`);
console.log(`Peak hour: ${data.peak_hour}`);
console.log(`Conversion rate: ${data.conversion_estimate}%`);

Response (200 OK)

{
  "footfall_id": "ftf_3m8k5n2p",
  "timestamp": "2026-02-11T18:00:00Z",
  "analysis_period": "2026-02-11T09:00:00Z to 2026-02-11T18:00:00Z",
  "total_visitors": 1842,
  "unique_visitors": 1653,
  "peak_hour": "14:00-15:00 (284 visitors)",
  "avg_dwell_time_minutes": 18.4,
  "zones": [
    {
      "name": "Entrance",
      "visitors": 1842,
      "avg_dwell_time_minutes": 2.1
    },
    {
      "name": "Electronics",
      "visitors": 628,
      "avg_dwell_time_minutes": 12.5,
      "conversion_estimate": "8.2%"
    },
    {
      "name": "Clothing",
      "visitors": 892,
      "avg_dwell_time_minutes": 24.7,
      "conversion_estimate": "12.4%"
    }
  ],
  "heatmap_url": "https://storage.chevora.ai/footfall/heatmap_feb11.png",
  "recommendations": [
    "Increase staff in Electronics (14:00-16:00)",
    "Optimize product placement in low-traffic northwest corner"
  ],
  "usage": {
    "video_hours_processed": 9.0,
    "cost": "€4.50"
  }
}

📊 Practical insights: optimize staffing, merchandising, and marketing based on real customer behavior.

Queue Management API

POSThttps://api.chevora.ai/v1/vision/queue

Estimate queue length and wait time in real time. Checkout optimization. EUR 550 per store per month or usage-based pricing.

Parameters

video_stream (file or URL, required) — Checkout camera stream
queue_zones (array, optional) — Queue zones to monitor [{"x": 100, "y": 200, "width": 400, "height": 300}]
alert_threshold (integer, optional) — People threshold for alert (default: 5)
webhook_url (string, optional) — URL for real-time alerts

Example: Python

response = requests.post(
    "https://api.chevora.ai/v1/vision/queue",
    headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
    json={
        "video_stream": "rtsp://camera1.store.com/checkout",
        "queue_zones": [
            {"x": 100, "y": 200, "width": 400, "height": 300, "name": "Checkout 1"},
            {"x": 600, "y": 200, "width": 400, "height": 300, "name": "Checkout 2"}
        ],
        "alert_threshold": 7,
        "webhook_url": "https://your-app.com/queue-alerts"
    }
)

queue_data = response.json()
print(f"Queue length: {queue_data['queue_length']}")
print(f"Estimated wait: {queue_data['estimated_wait_minutes']} min")

Response (200 OK)

{
  "queue_id": "queue_5k8m3n2p",
  "timestamp": "2026-02-11T14:35:22Z",
  "zones": [
    {
      "name": "Checkout 1",
      "queue_length": 8,
      "estimated_wait_minutes": 4.2,
      "alert_triggered": true
    },
    {
      "name": "Checkout 2",
      "queue_length": 3,
      "estimated_wait_minutes": 1.8,
      "alert_triggered": false
    }
  ],
  "total_customers": 11,
  "peak_hour_forecast": "17:00-19:00 (predicted wait: 6-8 min)",
  "staffing_recommendation": "Add 1 cashier to Checkout 1",
  "usage": {
    "video_minutes_processed": 5,
    "cost": "€0.055"
  }
}

💡 Business impact: retailers reduce checkout abandonment by 2-3%, translating to EUR 12K-18K annual revenue per midsize store.

Safety Monitoring API

POSThttps://api.chevora.ai/v1/vision/safety

Detect PPE (helmets, vests, gloves) and hazards. Site and warehouse safety. EUR 2,750 per site per month.

Parameters

video (file or stream, required) — Site camera video
detect_ppe (boolean, optional) — Enable PPE detection (default: true)
detect_hazards (boolean, optional) — Detect hazard zones (default: true)
compliance_mode (string, optional) — "osha" | "eu_directive" | "iso_45001"
alert_contacts (array, optional) — Email/SMS for immediate alerts

Example: cURL

curl -X POST https://api.chevora.ai/v1/vision/safety \
  -H "Authorization: Bearer chevora_sk_YOUR_KEY" \
  -F "video=@warehouse.mp4" \
  -F "detect_ppe=true" \
  -F "detect_hazards=true" \
  -F "compliance_mode=osha" \
  -F "alert_contacts[]=safety@company.com"

Example: Python (production)

import requests
from datetime import datetime

# Мониторинг в реальном времени с RTSP-камеры
response = requests.post(
    "https://api.chevora.ai/v1/vision/safety",
    headers={"Authorization": "Bearer chevora_sk_YOUR_KEY"},
    json={
        "video_stream": "rtsp://camera.construction-site.com/zone-a",
        "detect_ppe": True,
        "detect_hazards": True,
        "compliance_mode": "eu_directive",
        "alert_contacts": ["safety@company.com", "+421901234567"]
    }
)

safety_report = response.json()

# Обработка нарушений
for violation in safety_report['violations']:
    if violation['severity'] == 'critical':
        print(f"⚠️  CRITICAL: {violation['type']} at {violation['timestamp']}")
        print(f"   Worker ID: {violation['worker_id']}")
        print(f"   Location: {violation['coordinates']}")
        
# Журнал аудита для соответствия требованиям
with open(f"safety_audit_{datetime.now().date()}.json", "a") as log:
    log.write(json.dumps(safety_report) + "\n")

Response (200 OK)

{
  "safety_check_id": "safety_8x2m5k9w",
  "timestamp": "2026-02-11T10:22:13Z",
  "site": "Construction Site Alpha",
  "duration_minutes": 60,
  "workers_detected": 24,
  "violations": [
    {
      "type": "missing_helmet",
      "severity": "critical",
      "worker_id": "W-0142",
      "timestamp": "2026-02-11T10:35:08Z",
      "coordinates": {"x": 1240, "y": 680},
      "snapshot_url": "https://storage.chevora.ai/safety/snap_abc123.jpg",
      "alert_sent": true
    },
    {
      "type": "unauthorized_zone_entry",
      "severity": "high",
      "worker_id": "W-0089",
      "timestamp": "2026-02-11T10:48:22Z",
      "zone": "Crane Operation Zone",
      "coordinates": {"x": 880, "y": 420},
      "alert_sent": true
    }
  ],
  "compliance_score": 87.5,
  "compliance_standard": "EU Directive 89/656/EEC",
  "recommendations": [
    "Increase PPE enforcement in Zone A",
    "Review crane zone access protocols"
  ],
  "usage": {
    "video_hours_processed": 1.0,
    "cost": "€2.75"
  }
}

🚨 Compliance ready: audit logs meet OSHA, EU Directive 89/656/EEC, and ISO 45001 requirements. Reduce incident rates by 40-60%.

Storage

POST/v1/storage/upload

Upload documents for RAG and OCR pipelines.

file (binary, required) — File to upload
collection (string, required) — Collection name

Usage & Billing

Usage metrics are available in the dashboard and API. CSV exports and billing webhooks are supported.

SDKs & Libraries

Official SDKs for Python, Node.js, and CLI tools.

φ Python SDK

pip install chevora-ai

from chevora import ChevoraAI

client = ChevoraAI(
    api_key="chevora_sk_YOUR_KEY"
)

response = client.text.generate(
    prompt="Hello world",
    max_tokens=100
)

print(response.text)

⊛ Node.js SDK

npm install @chevora/ai-sdk

import { ChevoraAI } from '@chevora/ai-sdk';

const client = new ChevoraAI({
  apiKey: 'chevora_sk_YOUR_KEY'
});

const response = await client.text.generate({
  prompt: 'Hello world',
  maxTokens: 100
});

console.log(response.text);

⚡ CLI Tool

brew install chevora-cliornpm install -g @chevora/cli

# Configure API key
chevora configure --key chevora_sk_YOUR_KEY

# Text generation
chevora text generate "Write a summary" --max-tokens 200

# OCR extraction
chevora ocr extract invoice.pdf --output invoice.json

# Check usage
chevora usage --month 2026-02

📦 Postman Collection

Ready-to-use collection with all endpoints and examples.

Download Postman Collection

Webhooks Integration

Receive real-time event notifications.

Available events

invoice.processed

OCR completed for invoice

classification.completed

Classification completed

rag.query.answered

RAG query answered

usage.threshold.reached

Usage limit reached

api_key.rotated

API key rotated

invoice.payment_due

Invoice payment due

Webhook setup

POST https://api.chevora.ai/v1/webhooks

{
  "url": "https://your-app.com/webhooks/chevora",
  "events": ["invoice.processed", "usage.threshold.reached"],
  "secret": "your_webhook_secret"
}

Payload example

{
  "id": "evt_1a2b3c4d",
  "type": "invoice.processed",
  "created": 1707523800,
  "data": {
    "invoice_id": "ocr_9k3m2n1p",
    "invoice_number": "INV-2026-0042",
    "total_amount": 660.00,
    "currency": "EUR",
    "status": "completed"
  }
}

🔒 All webhook payloads are signed with HMAC-SHA256. Verify the signature for security.

Rate Limits

Default: 60 req/min per project. Dedicated limits and SLA available for enterprise.

Errors

400 — invalid parameters. 401 — missing key. 429 — rate limit. 500 — service error.