DaoXE vs official API cost, explained honestly

What you actually pay for#

With official APIs you pay each vendor's list price per 1M input/output tokens, plus the hidden overhead of running multiple accounts: separate cards, separate invoices, and — in many regions — failed checkouts. A gateway changes the shape of the bill: one key, one balance, per-model pricing, and the option to route cheaper models for cheap tasks.

Gateway vs official: the honest comparison#

How a DaoXE bill is assembled#

There is no plan, no seat count and no monthly minimum: you top up a balance and every request draws it down. Four things decide what a single call costs, and you can read all four before you spend anything. Top-up runs at a flat 1 RMB = $1 of credit on every payment method, and each model then bills at its own USD rate — live per-model rates are on the live pricing page.

Price your own workload in four steps#

1. Read your model's live rate. Open the live pricing page, find your exact model ID, and note the input and output rate for the group your key actually uses — not the cheapest group on the page.
2. Read the vendor's own list price. Open the model owner's official pricing page (linked below) for the same model, so you are comparing like with like rather than against a half-remembered figure.
3. Meter one real request. Send one small call at temperature=0, read usage.prompt_tokens and usage.completion_tokens from the response, then check the quota your account actually deducted.
4. Scale it to a real month. Multiply by your true monthly token volume — including retries, system prompts resent on every turn, and reasoning tokens. That is where cost estimates usually go wrong, not in the per-token rate.

Paying for it: methods and top-up#

You buy credit first and spend it per request; there is no subscription, and an unused balance simply stays a balance. Top-up runs at a flat 1 RMB = $1 of credit on every payment method, and each model then bills at its own USD rate — live per-model rates are on the live pricing page.

When a gateway saves money — and when it doesn't#

How the billing stays transparent#

• Per-model, per-token pricing published on the pricing page.
• Account-scoped model list, so the models you see are the ones you can call.
• Independent verification with the open benchmark — you don't have to take the numbers on faith.

Verify the cost advantage in one run#

The fair way to settle it: run the same fixed prompt through DaoXE and the official API at temperature=0, and compare token usage, latency and output. The benchmark guide automates that.

curl --fail-with-body --show-error --silent \
  https://daoxe.com/v1/chat/completions \
  -H "Authorization: Bearer ${DAOXE_API_KEY}" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "YOUR_EXACT_MODEL_ID",
    "max_tokens": 64,
    "messages": [{"role": "user", "content": "Say hello in one sentence."}]
  }'

Frequently asked questions#