Cut LLM costs: 400+ LLMs, 0% markup, pay in IDR.
One OpenAI-compatible endpoint for 400+ LLMs from 14+ providers. Tokens are sold at provider prices (0% markup), non-interactive workloads can be routed through batch, and billing is in Rupiah.
Where the savings come from
LLM costs balloon for three reasons: every task is sent to the most expensive model, background workloads are paid at real-time prices, and USD bills fluctuate with the exchange rate. Anoman solves all three from a single gateway — without you having to swap SDKs or rewrite integrations.
0% token markup
Token price = provider price. Anoman's revenue comes from a per-tier platform fee, not from quietly marking up your token prices.
400+ LLMs, 14+ providers
Switch from a premium model to a cheaper one for lightweight tasks (classification, extraction, summarization) by changing a single model string.
Batch routing
Workloads that don't need instant answers — document analysis, overnight reports, enrichment pipelines — are routed through the provider's batch API at a lower cost.
Billing in IDR
Pay in Rupiah — virtual account, QRIS, e-wallet, or card. No USD credit card, no unexpected exchange-rate gaps at month-end.
Pick a tier that matches your usage pattern
A best-effort guarded daily token bucket plus a signup credit. Enough to build a prototype and test guardrails before committing.
A fresh Rp 300rb of spendable credit each billing cycle across all model bands (Claude, GPT-4o included), with any overflow billed at raw provider cost. The monthly allowance resets on your renewal date — unused credit does not carry over. A predictable monthly plan for production apps.
Buy a Budget or Plus day/week pass for a guaranteed daily token allowance — no subscription. Ideal for bursty or seasonal usage, or top up a prepaid balance at raw provider cost (0% token markup) when it rises and falls.
A full comparison of all tiers (including Enterprise) is on the pricing page.
Per-request cost transparency
Every response carries an _anoman block with the actual cost in USD — so you know exactly what each call costs, instead of waiting for the month-end bill.
# Expensive and cheap models through the same endpoint —
# change the model string, not the integration.
curl https://api.anoman.io/v1/chat/completions \
-H "Authorization: Bearer anm-sk-..." \
-d '{
"model": "gpt-4o-mini",
"messages": [{"role": "user", "content": "Classify the sentiment of this review..."}]
}'Bonus: these savings don't come at the expense of security — guardrails (PII masking, injection detection, content moderation) still run on every request. Related: Indonesian-language chatbot with guardrails and UU PDP readiness for AI apps.
Frequently asked questions
What does 0% token markup mean?
The token price you pay equals the provider's published price — Anoman does not mark up token prices. Anoman earns from its plans instead: a free guarded daily tier, day/week Passes from Rp 10rb, or Pro at Rp 499rb/month with Rp 300rb of spendable credit. Prepaid balance top-ups (raw provider cost, no monthly fee — a small platform fee applies at top-up) are available on any plan.
How does batch routing save on LLM costs?
Workloads that don't need instant answers — document analysis, overnight reports, enrichment pipelines — are routed through the provider's batch API, which is priced lower than real-time. You send the request as usual, receive a 202 with a job ID, then fetch the result by polling; the _anoman block lists the savings.
Can I pay in Rupiah?
Yes. Anoman billing is available in IDR and can be paid via virtual account, QRIS, e-wallet, or card. No USD credit card, no unexpected exchange-rate gaps at month-end.