Start with RM5

AI API for Malaysia. One key. From RM5.

GPT, Claude, DeepSeek and more with prepaid MYR credits. FPX and DuitNow. No subscription.

Start with RM5See available models
from openai import OpenAI
client = OpenAI(api_key="asai_sk_...", base_url="https://serveras.click/v1")
client.chat.completions.create(model="asai/gpt-6-sol",
    messages=[{"role": "user", "content": "Hello"}])
OpenAI-compatible setup guide
54 available modelslive catalog
4 free modelsno asAi credit charge
1.05M max contextlargest current window
1 prepaid balanceshared across models

Pay in MYR, not USD.

asAi sells prepaid credit in Malaysian ringgit, so a Malaysian card never meets a foreign charge or monthly invoice. Every pack opens the same available model catalog; bigger packs only add more credits per ringgit, never extra model tiers. There is no subscription to cancel and no card stored after the FPX or DuitNow approval completes on your phone.

Choose a pack that matches your workload, review the current input, cache and output multipliers before sending traffic, then watch each deduction land in Logs against the same balance. Validity is visible in the dashboard from the redemption timestamp.

Current prepaid packs in MYR
PriceCreditsCredits per RMValidity
RM580,000,00016M720 hours
RM8140,000,00017.5M720 hours
RM10200,000,00020M720 hours

Pack credits are deliberately non-linear. A larger pack raises the credits received per RM without changing model access or multipliers. Start with the RM5 pack to validate integration, then move up only when request volume requires it.

Deploy anywhere.

The asAi API runs behind Cloudflare, offering low latency routing for Malaysian endpoints while remaining fully accessible from any global region. Teams can deploy serverless functions, Edge Workers, containers on AWS or local servers in Kuala Lumpur without encountering region locks or requiring separate keys per environment.

Unlike rigid subscription models that require seat licensing or limit concurrency artificially, asAi provides 7 concurrent requests and 20 requests per minute across all accounts by default. This baseline supports continuous agentic workflows, concurrent data extraction across small worker fleets, and parallel chat streams without unexpected rate-limit exhaustion.

Start with small experiments.

The RM5 pack is explicitly designed for prototyping and testing. By providing 80,000,000 credits without an ongoing commitment, teams can build a proof of concept, validate prompt performance across Kimi, DeepSeek and Claude, and present a working prototype before engaging corporate procurement.

There are no surprise charges. Because asAi holds no stored card and operates purely on a prepaid balance, a runaway loop or an accidental high-context request will safely exhaust the RM5 pack and stop, rather than generating an unpayable invoice. The system enforces safety by design.

One balance for every model.

The same prepaid balance funds text, vision and long-context requests across the catalog. Pick a different alias per task without opening another account, moving credits or accepting a tier lock. Confirm each alias in Models before depending on it, because aliases retire and new ones appear.

Free models do not deduct asAi credits, while paid models apply their published multipliers to the shared balance. Cache-friendly prompts can lower the input cost when the selected model supports prompt caching.

Built for OpenAI-compatible tools.

Point the OpenAI SDK at https://serveras.click/v1, keep your existing request code, and pick any current model alias. The same key works for chat completions, Hermes Agent, Claude Code and Codex when each client uses its documented compatibility setup. The full two-value switch is covered in the OpenAI-compatible guide.

curl https://serveras.click/v1/models

What RM5 buys.

RM5 pack gives 80,000,000 credits for 720 hours. No subscription. Every available model shares the same balance at its published input, cache and output rates.

  1. Illustrative example A: 200 input + 650 output tokens on asai/gpt-6-sol = 34,500 credits, about RM0.00216, about 2,318 similar requests per RM5 pack.
  2. Illustrative example B: 10,000 input + 1,000 output tokens on asai/claude-opus-5 = 225,000 credits, about RM0.01406, about 355 similar requests per RM5 pack.
  3. Illustrative example C: 100 input + 1,500 output tokens on asai/deepseek-v4-pro = 38,000 credits, about RM0.00238, about 2,105 similar requests per RM5 pack.

Recheck exact multipliers in Billing before you publish cost claims. Actual deduction depends on cache and generated output.

How to start in three steps.

  1. 01
    Create an account.

    Sign up with Google, verify your email, and open the dashboard.

  2. 02
    Top up RM5 with FPX or DuitNow.

    Choose the starter pack, approve the bank step, and confirm the balance.

  3. 03
    Create a key and send a request.

    Make a key under API keys, set the base URL, and run the example above.

AI API Malaysia FAQ.

Is the RM5 pack a subscription?

No. It is a one-time prepaid pack. There is no auto-renewal and no card on file.

How long do credits last?

Credits stay valid for 720 hours from top-up. The timer starts when the pack is redeemed, not when you register.

How do I pay with FPX or DuitNow?

Open Top up, choose a pack from RM5, and finish the bank approval on your phone.

Can the RM5 pack use every model?

Yes. Pack size does not gate model access. Pick any alias in Models; availability and account controls still apply.

Where do I see credit usage?

Open Logs after a request. It shows the model, token counts, timestamp and credits deducted.

Prefer an agent workflow? Read Setup Hermes Agent di Malaysia dengan asAi or the English technical guide.

Start with RM5