Route your AI
requests with precision
Access multiple AI models through one unified API. Pay only for what you use with transparent token-based billing powered by eSewa.
Get started in seconds
Sample code. Use model docs for the exact curl command.
Terminal
$ curl https://api.corerouter.me/v1/complete \
-H "Authorization: Bearer pk_test_abc123..." \
-H "Content-Type: application/json" \
-d '{
"model": "mistral-7b",
"messages": [{
"role": "user",
"content": "Explain quantum computing"
}]
}'pk_test_project_a123...
API Key A
Can be used with any model in CoreRouter
Daily: 50 / 120
Monthly: 360 / 1000
pk_test_project_b456...
API Key B
Can be used with any model in CoreRouter
Daily: 28 / 80
Monthly: 190 / 700
pk_test_project_c789...
API Key C
Can be used with any model in CoreRouter
Daily: 16 / 60
Monthly: 120 / 500
Create Multiple API Keys
Organize your integrations with separate API keys for teams, projects, or environments. Any API key can be used with any model available in CoreRouter.
- Any key works with any model in the system
- Set daily request limits per key
- Configure monthly usage quotas
- Easy rotation and revocation
Features
Everything you need to integrate AI into your application
Multi-Model Support
LLMs, OCR, and custom models. Switch providers without changing your code.
Token-Based Pricing
Pay for exactly what you use. No hidden fees, no monthly minimums.
eSewa Payments
Instant top-ups built for Nepal. Secure and convenient.
Simple API
RESTful API with clear documentation and SDKs for popular languages.
Real-Time Analytics
Track usage, costs, and performance with granular insights.
Enterprise Security
API keys, rate limiting, and role-based access control.
Available Models
Choose from our growing catalog of AI models
Qwen3.5-Flash
by Qwen
A fast and cost-effective language model optimized for real-time applications. It delivers quick responses with reasonable accuracy, making it ideal for chat applications, autocomplete systems, and high-throughput API usage.
ChatGPT-4.1
by ChatGPT
A high-performance large language model designed for advanced reasoning, content generation, and conversational AI. It excels at complex problem-solving, coding assistance, and structured outputs, making it ideal for production-grade AI applications.
Gemini-3.1-Pro
by Gemini
A robust multimodal AI model capable of processing both text and images with high accuracy. It performs well in OCR, visual reasoning, and contextual understanding, making it ideal for applications requiring image-to-text and intelligent analysis.
Mistral-7B
by Mistral
A lightweight yet efficient open-weight language model optimized for fast inference and low-cost deployments. It is suitable for chatbots, text generation, and API-based services where performance and cost efficiency are critical.
Qwen/Qwen2-0.5B-Instruct
by Qwen
A compact instruction-tuned model designed for lightweight tasks and efficient deployments. It performs basic OCR and text understanding tasks while maintaining low resource usage, making it ideal for edge or budget-constrained environments.
Claude Opus 4.6
by Claude
A powerful multimodal model optimized for document understanding and OCR tasks. It can accurately extract, interpret, and summarize text from images and complex documents, making it suitable for enterprise document processing workflows.
Transparent Pricing
See what our most popular models cost
| Model | Type | Pricing |
|---|---|---|
| Qwen3.5-Flash | LLM | — |
| ChatGPT-4.1 | LLM | रू 0.0025/1k input · रू 0.01/1k output |
| Gemini-3.1-Pro | OCR | — |
| Mistral-7B | LLM | रू 0.003/1k input · रू 0.015/1k output |