What the OpenAI API costs. And the same models for 20 to 25% less.
GPT prices per million tokens, from OpenAI's pricing page, next to what the same model IDs cost on APIsRouter. Keep the OpenAI SDK and change one base URL.
The same traffic for 30 days
Tokens per request: support 3,000 in (2,000 cached) and 500 out; summaries 1,000 in and 100 out; agent 68,000 in (60,000 cached) and 3,000 out.
| OpenAI | APIsRouter | |
|---|---|---|
| Support bot, 1,000 answers a day (GPT-5.5) | $630 | $504 |
| Summaries, 10,000 documents a day (GPT-5.6 Luna) | $96 | $74 |
| Coding agent, 200 tasks a day (GPT-6 Astra) | $1,740 | $1,392 |
| Total | $2,466 | $1,970 |
OpenAI API prices, per million tokens
Standard tier. Where APIsRouter sells a model below list, both prices are shown. OpenAI prices from developers.openai.com/api/docs/pricing, September 2026. Standard tier, prompts up to 272,000 tokens.
| Model | Input: OpenAI / APIsRouter | Cached input: OpenAI / APIsRouter | Output: OpenAI / APIsRouter |
|---|---|---|---|
| GPT-6 Astra | $10 / $8 | $1 / $0.80 | $50 / $40 |
| GPT-6.1 Sol | $2 / $1.50 | $0.10 / $0.08 | $10 / $8 |
| GPT-6 Sol | $2 / $1.50 | $0.20 / $0.15 | $10 / $8 |
| GPT-6 Luna | $0.10 / $0.08 | $0.01 / $0.008 | $0.50 / $0.40 |
| GPT-5.6 Sol | $4 / $3 | $0.40 / $0.30 | $20 / $16 |
| GPT-5.6 Terra | $2 / $1.50 | $0.20 / $0.15 | $12 / $9.50 |
| GPT-5.6 Luna | $0.20 / $0.15 | $0.02 / $0.015 | $1.20 / $0.95 |
| GPT-5.5 | $5 / $4 | $0.50 / $0.40 | $30 / $24 |
| GPT-5.5 Pro | $30 / Official price only | — | $180 / Official price only |
| GPT-5.4 | $2.50 / Official price only | $0.25 / Official price only | $15 / Official price only |
| GPT-5.4 mini | $0.75 / Official price only | $0.075 / Official price only | $4.50 / Official price only |
How OpenAI bills a request
- Cached input is a fraction of the price: Prompt prefixes OpenAI has seen recently bill as cached input: $0.50 instead of $5 per million on GPT-5.5. Keeping system prompts and tool lists stable keeps them cached.
- Output costs 5 to 6 times input: Every model on the list prices output well above input, so long answers drive the bill more than long questions.
- Reasoning tokens bill as output: Models that think before they answer bill those hidden reasoning tokens as output, so a short reply can still use many output tokens.
- Long prompts cost more: Above 272,000 input tokens, models with long-context pricing charge double for input and 1.5 times for output.
Switch with one line
- Get a key: Top up from $10 below. The key and a login link arrive by email.
- Point the OpenAI SDK at APIsRouter: Set base_url to https://api.apisrouter.com/v1. Model IDs stay the same.
- Send requests: Every request shows up in your console with its tokens and cost.
from openai import OpenAI
client = OpenAI(
base_url="https://api.apisrouter.com/v1",
api_key="sk-your-apisrouter-key",
)
reply = client.chat.completions.create(
model="gpt-5.5",
messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)Two ways to pay
Both fill one balance, spent at the prices above. Pay as you go from $10 with no account, with +30% on the first top-up.
FAQ
How much does the OpenAI API cost?
From $0.10 / $0.50 per million input and output tokens on GPT-6 Luna to $30 / $180 on GPT-5.5 Pro. GPT-5.5 costs $5 / $30 on OpenAI and $4 / $24 on APIsRouter.
Can I try it for free?
Yes. Sign in with your email or Google and claim a $2 trial credit from your dashboard, no card needed. When it runs out, top up from $10.
What is the cheapest OpenAI model?
GPT-6 Luna, at $0.10 / $0.50 per million input and output tokens on OpenAI. The cheapest OpenAI model on APIsRouter is GPT-6 Luna at $0.08 / $0.40.
How is cached input billed?
Input that repeats a recent prompt prefix bills at the cached rate, $0.50 instead of $5 per million on GPT-5.5. APIsRouter bills cached input the same way, at its own lower rate.
Do reasoning tokens cost extra?
They are billed as output tokens, at the output price, and each response reports them in its usage.
Can I keep using the OpenAI SDK?
Yes. Set base_url (baseURL in Node) to https://api.apisrouter.com/v1 and use your APIsRouter key. The model IDs are the same, and streaming and tool calls work as before.
Is there a subscription or a minimum?
No subscription. Top up from $10 and pay per token, with +30% on the first top-up. Credit never expires.