About LLM Resayil

The LLM API Built for
Developers Who Ship

45+ open-source and frontier AI models through one OpenAI-compatible gateway. Local GPU inference, cloud proxies, pay-per-token — zero lock-in.

Get Started Read the Docs
45+
AI Models
1:1
OpenAI Compatible
KWD
Local Currency
0ms
Setup Time
Our Mission

Powerful AI shouldn't require
a PhD in infrastructure

We built LLM Resayil to remove the friction between "I want to use AI" and "my app is live." GPU servers, model weights, provider APIs, billing — we handle it. You focus on building.

What We Offer
OpenAI-Compatible API
Drop-in replacement. Change your base URL, keep your code. Works with any SDK that targets OpenAI's API — Python, Node, curl, n8n, LangChain.
$ curl https://llm.resayil.io/api/v1/chat/completions
  -H "Authorization: Bearer $API_KEY"
  -d '{"model":"llama3.2:3b","messages":[...]}'
# Returns standard OpenAI-format JSON ✓
45+
AI Models
Local GPU: Llama 3.2, Qwen 3, Mistral, Gemma, Phi-4. Cloud proxies: DeepSeek V3.1, GPT-4o, Claude, Gemini, and more.
15 local · 30 cloud
Pay Per Token
No subscriptions. Buy credits, use them anytime. Local: 1 cr/token. Cloud: 2 cr/token.
No monthly fee
KNET Payments
Secure local KNET payments via MyFatoorah. KWD pricing. No international card needed.
Powered by MyFatoorah
Full Usage Visibility
Per-call logs: model, tokens, credits, timing. Weekly summaries in your dashboard.
Secure by Default
API keys hashed at rest. WhatsApp OTP for account verification. Per-key rate limits. HTTPS only.
Pricing

Simple credit-based pricing

Credit top-ups — unused credits never expire
Credits
Price (KWD)
Local use
Cloud use
5,000 credits
2.000 KWD
5,000 tokens
2,500 tokens
15,000 credits
5.000 KWD
15,000 tokens
7,500 tokens
50,000 credits
15.000 KWD
50,000 tokens
25,000 tokens
Infrastructure

Built on dedicated GPU hardware

We run our own GPU server for local model inference, backed by cloud proxy routing for frontier models. Everything is served over HTTPS with per-key authentication.

Local GPU Server

Dedicated GPU hardware running Ollama for fast, private inference on open-source models. Llama, Qwen, Mistral, Phi, Gemma, and more — served from our own hardware.

Cloud Proxy Models

Frontier models (DeepSeek V3.1, Qwen 3.5 397B, and others) routed via Ollama cloud proxies. Same API interface — no extra configuration.

Payment Processing

All payments handled by MyFatoorah — a licensed Kuwaiti payment gateway. We never store card numbers or KNET credentials.

Account Security

Phone verification via WhatsApp OTP. API keys are hashed, never stored in plain text. Rate limiting applied per key, per tier.

Ready to build?

Create your account, get your API key, and make your first request in under 5 minutes.

Create Free Account Explore the Docs