Transparent Developer Pricing
Predictable Cost. Generous Free Tier.
Start prototyping with 16,666 free requests every day on RapidAPI. No credit card required. Scale up to millions of decisions at a fraction of generative LLM pricing.
Active Now • Free Launch
Developer Launch Tier
Developer Free
Complete access for hobbyists, production prototypes, and workflow automation.
$0
/ month
16,666 calls / day
(Up to ~500,000 calls / month free)
- ✓ All 6 core decision endpoints unlocked
- ✓ No credit card or payment info required
- ✓ Sub-35ms raw neural tensor pass
- ✓ 5 requests / second burst concurrency
- ✓ Strict zero-retention RAM processing
- ✓ Community Discord & GitHub support
Coming Soon
High Concurrency
Laya Pro
Dedicated cloud capacity for high-volume SaaS backends and high-traffic AI firewalls.
~$19
/ month (Targeted)
1,000,000 calls / month
Tier in active preparation • Waitlist open
- ✓ Priority cloud worker execution queue
- ✓ 100 requests / second throughput rate
- ✓ 99.9% uptime SLA target with alerts
- ✓ Automatic burst overage allocation
- ✓ Direct priority email support from Harshad
Enterprise / Private Cloud
Self-Hosted Container
Run Laya inside your own private VPC or air-gapped infrastructure. Zero data egress.
Custom
/ annual license
Unlimited In-VPC Calls
Zero per-request fees • 0ms WAN transit
- ✓ Deploy on Docker, Kubernetes, OpenShift, or AWS
- ✓ Infisical & HashiCorp Vault secrets integration
- ✓ 100% data residency & HIPAA / GDPR airgap compliance
- ✓ Custom categorization taxonomy tuning
- ✓ Dedicated architecture review with Harshad Jadav
Real Numbers: Laya vs. Generative LLMs
Comparing the actual costs and latency of processing 1,000,000 customer decisions (support ticket triage, spam classification, or prompt injection validation):
| Engine | Monthly Cost (1M Calls) | Inference Latency | Hallucination Risk | Prompt Injection Safe |
|---|---|---|---|---|
| Laya Decision Gateway (Free Launch / Planned Pro) | $0 (Free Launch) / ~$19 (Planned Pro) | < 35ms core | 0% (Deterministic) | ✓ Hardware-calibrated |
| GPT-4o mini (OpenAI) | ~$75.00 | ~1,200ms | ~2 – 4% | ✗ Vulnerable to jailbreak |
| Claude 3.5 Haiku (Anthropic) | ~$100.00 | ~1,400ms | ~2 – 3% | ✗ Vulnerable to jailbreak |
| GPT-4o (Full) | ~$1,250.00 | ~2,500ms | ~1 – 2% | ✗ Vulnerable to jailbreak |
* Calculation assumes an average input of 150 tokens and structured classification output of 25 tokens per request. During the current public launch period, up to 16,666 calls/day (~500k/month) are completely free with zero billing setup.
Frequently Asked Questions
Do I really get 16,666 calls every day for free without a credit card?
Yes. Through our RapidAPI launch tier, developers receive 16,666 free requests per calendar day (which resets at 00:00 UTC). RapidAPI does not require any payment method or credit card to subscribe to this tier.
What happens if my application exceeds the daily free quota?
On the free tier, requests beyond 16,666 in a single 24-hour window will receive an HTTP 429 (Rate Limit Exceeded) status until the midnight UTC reset. We are actively preparing the Pro tier for unlimited burst overages. If your production workload requires higher limits before Pro is live, email Harshad directly at harshadjadav@duck.com or consider our self-hosted container option.
Can I deploy Laya inside my company's AWS, GCP, or Azure VPC?
Yes. With an Enterprise container license, you receive a hardened Docker container image that boots into memory in under 20 seconds. It has zero external dependencies, requires no GPU, and does not make any outbound telemetry calls, guaranteeing strict air-gapped security and HIPAA/GDPR compliance.