OpenAI API Pricing, Models & Rate Limits (2026)
OpenAI trains and serves the GPT model family, the API whose request/response shape every other provider on this site now imitates. The GPT-5.6 line is the current flagship generation, spanning a reasoning-heavy top tier down to a cheap, low-latency tier built for high-volume calls.
Models and prices
| Model | Input /M | Output /M | Blended /M |
|---|---|---|---|
| GPT-5.6 Luna | $0.20 | $1.20 | $0.45 |
| GPT-5.6 Terra | $2.00 | $12.00 | $4.50 |
| GPT-5.6 Sol | $5.00 | $30.00 | $11.25 |
12 legacy OpenAI models
| GPT-5 Nano | $0.14/M blended |
| GPT-4o Mini | $0.26/M blended |
| GPT-5.4 Nano | $0.46/M blended |
| GPT-5 Mini | $0.69/M blended |
| GPT-5.4 Mini | $1.69/M blended |
| o3-Mini | $1.93/M blended |
| GPT-5 | $3.44/M blended |
| GPT-4.1 | $3.50/M blended |
| GPT-4o | $4.38/M blended |
| GPT-5.4 | $5.63/M blended |
| GPT-4 Turbo | $15.00/M blended |
| GPT-5.4 Pro | $67.50/M blended |
Speed
Fastest measured OpenAI model is GPT-5.6 Luna at 126 tokens/sec (300ms TTFT), median across measured OpenAI models is 78 tokens/sec. See the full speed benchmark methodology.
Operational facts
Verified 2026-08-08 against source.
| OpenAI-compatible | Yes |
| API base URL | https://api.openai.com/v1 |
| Auth model | Bearer API key |
| Prompt caching | Yes |
| Batch discount | 50% |
| Free tier | None |
| Rate-limit model | Usage tiers 1-5, auto-promoted by cumulative spend and account age |
| Data residency | US by default; EU data residency available on enterprise agreements |
| Trains on API data | No |
| SLA published | Yes |
Lifecycle
OpenAI has 12 legacy models still routable. Full dates and successors on the model deprecation tracker.
Switching to and from OpenAI
Calling OpenAI through All AI Ask
curl https://allaiask.com/api/v1/chat \
-H "Authorization: Bearer $ALLAIASK_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "gpt-5.6-luna", "messages": [{"role": "user", "content": "Hello"}]}'FAQ
Is OpenAI OpenAI-compatible?
Yes — OpenAI's API base (https://api.openai.com/v1) accepts the OpenAI SDK request/response shape, so existing OpenAI client code works with a base-URL and key swap.
Does OpenAI support prompt caching?
Yes, as of 2026-08-08 — see https://platform.openai.com/docs for the current mechanics and discount.
Does OpenAI have a free tier?
No published free tier as of 2026-08-08 — billing starts from the first request.
How much does the OpenAI API cost?
Current OpenAI models range from $0.45 to $11.25 per million blended tokens (3:1 input:output). Full per-model pricing is in the table below.
Where is OpenAI API data hosted?
US by default; EU data residency available on enterprise agreements