Credits and billing
aixplain is prepaid: each workspace buys credits, and every model, tool and agent run draws from that balance, at 1 credit = 1 USD.
Each workspace has its own balance. Credits cannot move between workspaces; see Workspaces.
What you pay for
| Asset | Priced by | Unit | aixplain fee |
|---|---|---|---|
| Model | The model provider | Usually tokens for LLMs; audio minutes, images or characters for other models | None on direct calls |
| Tool | The tool provider | Usually per request | None on direct calls |
| Integration | Not billed itself | The tools you create from it are billed as tools | None |
| Agent run | Sum of every model and tool call in the run | Per run | 20% on top of that sum |
The 20% applies only to calls made inside an agent run; it covers orchestration, inspectors, memory and tracing. For example, an agent run that makes 0.50 credits of LLM calls and 0.10 credits of tool calls costs 0.60 + 0.12 = 0.72 credits.
Current plans and prices are at aixplain.com/pricing. Each model's and tool's price is on its card in the Marketplace.
Add credits
Owners and Admins add credits in the aixplain app at Settings > Billing.
- Open Settings > Billing in the workspace you want to fund.
- Click Add credits.
- Enter the amount, choose a payment card, and confirm.
To use a voucher code, open the same page, choose to redeem a voucher and enter the code. Voucher credits are added to the balance at once.
When the balance reaches 0, runs that cost credits are refused with the error code err.not_enough_balance until you add credits.
Track spending
Every run reports its cost, in credits, in result.used_credits.
from aixplain import Aixplain
aix = Aixplain(api_key="YOUR_API_KEY")
model = aix.Model.get("openai/gpt-4o-mini/openai")
result = model.run(text="Name one primary color.")
print(result.data)
print(result.used_credits)
For the full history, open Settings > Usage. It lists each transaction with its date, asset, data volume and credits.
To cap spending, set a budget on an agent (Budgets) or limits on an API key (Rate limiting).
Next steps
- Rate limiting: cap the credits and requests each API key can use.
- Budgets: cap the cost of a single agent run.