nepal·ai·gateway

API live · checking seats

One key.
Every frontier model.

Claude Opus 4, GPT 5.5, Kimi, DeepSeek, MiniMax, Gemini and more — through a single OpenAI-compatible endpoint. No global cards, no per-provider signups. Built to be reached from Nepal.

Make your first request See the plan

Works with the OpenAI SDKs and tools you already use.

SEATS —
DAILY BUDGET 200K/day
RATE 15 RPM
STATUS live status loading

// model catalog

Pick the model. Change one line.

Now serving Claude Opus & Sonnet 4 🔥 — alongside fast, coding and long-context models. Same endpoint for all of them.

🔥 NEW · FLAGSHIP CLAUDE

Claude, on your key.

The models people ask for by name. Top-tier reasoning, writing, and agent work — live now.

claude-opus-4-8 claude-sonnet-4-6 kimi-2.6
FLAGSHIP

Strongest general choices.

Hard reasoning, polished writing, planning and research.

gpt-5.5mimo-v2.5-pro
CODING

Build, debug, understand code.

Focused coding models for features, reviews and long technical tasks.

kimi-k2.7-codedeepseek-v4-pro
FAST

Quick without giving up quality.

Chat, extraction, summaries and everyday app features.

gemini-3.5-flashdeepseek-v4-flash
LONG CONTEXT

Keep more in one conversation.

Large documents, long notes and multi-step project context.

minimax-m3nemotron-3-ultra

Your account is the source of truth for enabled models — or call GET /v1/models before you choose.

// why this exists

The account problem, solved.

01

No global card needed

Getting an OpenAI or Anthropic account from Nepal is a wall. One key here gets you past it.

02

One bill, many models

Stop juggling five provider dashboards. A single plan, a single key, every model.

03

Honest limits

We tell you what burns quota — Claude models cost ~4× per token. Clear errors, no silent charges.

04

A real person

Local support when you need it. Ask before you pay; we confirm seats first.

// from zero to first response

Three steps. Familiar API.

  1. 1

    Request access

    Join the support group and ask for the current plan. We confirm availability before you pay.

  2. 2

    Get your key

    After payment is confirmed, your key is issued to you and shown once.

  3. 3

    Point & build

    Drop the base URL and key into any OpenAI-compatible SDK. Done.

// developer access

Try every model for 8 days.

One short plan — enough to ship a prototype or finish a project without opening accounts everywhere.

Ask for access Usage guide

Payment is confirmed manually. Ask support for current eSewa details before sending.

NPR599/ 8 days
Daily allowance
200K weighted tokens
Total allowance
1.6M weighted tokens
Request speed
up to 15 / min
Models
top + fast selection
Rollout
25 seats total

Claude models draw quota ~4× faster per token — pick the model that fits the task.

// already have a key?

Know exactly where you stand.

Sign in with your key to see models, quota, rate limit and expiry — without revealing the full key again.

Check my account

// good to know

Straight answers.

Who is this for?

Students, developers and small teams who want several frontier models through one familiar API.

Do I change my OpenAI code?

Usually just the base URL, key and model name. Docs include cURL, Node.js and Python.

Which models will my key include?

Your exact list appears in your account after activation. The account page and GET /v1/models are authoritative.

What happens at a limit?

The API returns a clear error instead of quietly charging more. Your account shows remaining quota and expiry.

What happens after 8 days?

The key stops working when the window ends. Contact support for a renewal review.