# 3 Meta models · example with llama-4-maverick
import os
from openai import OpenAI
client = OpenAI(base_url="https://ai.helipod.io/v1",
api_key=os.environ["HELIPOD_AI_KEY"])
r = client.chat.completions.create(
model="llama-4-maverick",
messages=[{"role": "user", "content": "Halo!"}])Llama API pricing
Final prices in Rupiah per 1 million tokens. Cached = prompt tokens the provider already has in its cache (repeated system prompts, long documents).
Cost calculator
Per request
| Fresh input | 1,000 × Rp 8.900 | Rp 8,9 |
| Cached input | 1,000 × Rp 8.900 | Rp 8,9 |
| Output | 500 × Rp 26.700 | Rp 13,35 |
| One request | Rp 31,15 | |
A month (500 requests/day × 30)
Rp 467.250
An estimate: you are billed per request for the tokens actually used.
Llama API
The Llama API is programmatic access to the Llama models by Meta, a widely used family of open-weight models. Helipod offers Llama 4 Maverick, Llama 3.3 70B Instruct and Llama 3.1 8B Instruct.
You don't need to rent GPUs or run the models yourself: call Llama with the OpenAI or Anthropic SDK and one Helipod API key. Usage is billed per token in Rupiah from your Helipod balance.
A separate account, dollar billing and an international credit card.
The same Meta models, billed in Rupiah from one balance.
Advantages
Comparison
| Direct from Meta | Helipod AI Gateway | |
|---|---|---|
| Billing currency | Dollars (USD) | Rupiah, per token |
| Payment | International credit card | QRIS / virtual account |
| Other models | Meta models only | Claude, GPT, Gemini, DeepSeek, etc. |
| SDK | Meta SDK | OpenAI & Anthropic SDKs |
| Spend cap per key | Per organization | Per API key, monthly |
| Request history | Meta dashboard | Helipod console, prices recorded |
The Helipod team is happy to help you pick the right service.
Use cases
FAQ
Other models on the Helipod AI Gateway, with the same API key
Compare every model in the AI Gateway price list.