Provider pricing hub
OpenAI API pricing at the catalog rate.
As of 18 August 2026, RouterPlex lists 6 OpenAI routes — 5 chat models and 1 image-generation route. Token prices are vendor list rates with 0% markup. The calculator uses those live numbers, not a scraped marketing table.
6
Routes
$1.00
Cheapest input / 1M
OpenAI pricing calculator
Same workload. Every GPT route.
Defaults are a 10K-in / 1K-out request, 3,000 times a month. Switch to the 100K / 10K agent turn if you are costing a coding loop. Image-generation is listed below the table, not inside the estimate.
Workload calculator
Catalog rates · USD · no markup
Models in the estimate
4 selected| Route | Per request | Monthly | $100 buys |
|---|---|---|---|
| GPT-5.6 Luna OpenAI | $0.016 | $48.00 | 6,250 req |
| GPT-5.4 OpenAI | $0.040 | $120.00 | 2,500 req |
| GPT-5.6 Terra OpenAI | $0.040 | $120.00 | 2,500 req |
| GPT-5.5 OpenAI | $0.080 | $240.00 | 1,250 req |
Estimates apply live catalog input and output rates only. Actual spend follows the tokens your client sends and the model returns. Image-generation routes are omitted because they are not a chat turn.
OpenAI family
Published catalog rates.
Agent-turn cost uses 100K input and 10K output so Luna cannot hide behind a cheap input rate.
| Model route | Context | Input / 1M | Output / 1M | Agent turn |
|---|---|---|---|---|
GPT-5.6 Luna gpt-5.6-luna | 258K | $1.00 | $6.00 | $0.16 |
GPT-5.4 gpt-5.4 | 1M | $2.50 | $15.00 | $0.40 |
GPT-5.6 Terra gpt-5.6-terra | 258K | $2.50 | $15.00 | $0.40 |
GPT-5.5 gpt-5.5 | 256K | $5.00 | $30.00 | $0.80 |
GPT-5.6 Sol gpt-5.6-sol | 258K | $5.00 | $30.00 | $0.80 |
How to pick
Price is the easy part. Context is not.
These notes only use catalog fields: price, context, and mode. They are not quality rankings.
Cheapest chat route
GPT-5.6 Luna is the low-cost OpenAI default in this catalog. Use it when the task is short and you will retry rather than spend frontier rates on a first draft.
Same price, different window
GPT-5.4 and GPT-5.6 Terra currently share the same $2.50 / $15 rates. 5.4 is the one that publishes a 1M context window. If you are paying mid-tier rates, check the window before you pick on name.
One key, one ID change
Moving from Luna to Sol, or from OpenAI to Claude, is a model-string change on the same OpenAI-compatible endpoint. Per-key budgets still apply.
Compare the rest of the catalog on all model pricing or the cheapest-first ledger. Quality evidence lives on benchmarks.
Before you bill production
OpenAI pricing, in questions.
01How much does the OpenAI API cost on RouterPlex?+
As of 18 August 2026, RouterPlex bills OpenAI chat models at the live catalog rate with 0% markup. GPT-5.6 Luna is the cheapest chat route at $1.00 per 1M input tokens and $6.00 per 1M output tokens. The same rates appear on each model page and in the calculator above.
02What is the cheapest OpenAI model here?+
GPT-5.6 Luna is the cheapest OpenAI chat route in this catalog: $1.00 / $6.00 per 1M tokens, about $0.16 for a 100K-in / 10K-out agent turn.
03Does RouterPlex mark up OpenAI token prices?+
No. OpenAI chat and image routes are billed at the vendor list price shown in the catalog. There is no per-token markup and no fee when you top up. You pay for the tokens the model actually uses, from a prepaid balance that stops at $0.
04Can I use these models with the OpenAI SDK?+
Yes. Point any OpenAI SDK or tool at https://api.routerplex.com/v1, use your RouterPlex key, and set the model string to the catalog id (for example "gpt-5.6-luna"). The request shape stays Chat Completions.
05Why might a number here differ from openai.com/pricing?+
This page shows the rate RouterPlex bills today, pulled from the live catalog on 18 August 2026. Official marketing pages can change, hydrate in JavaScript, or list a promotional cut we have not been charged yet. If the two disagree, the catalog rate is what a request on this key costs.
06Which OpenAI model has the largest context window?+
GPT-5.4 publishes the largest context in this OpenAI set (1M). The other chat routes sit around 256K–258K. Context is a catalog field, not a quality score.
07Is GPT Image 2 billed like chat models?+
GPT Image 2 is an image-generation route. RouterPlex still shows input and output rates per 1M tokens, but a generated image is not comparable to a chat turn, so it is excluded from the calculator and from cheapest-model rankings.
Route OpenAI from one balance
One key reaches every GPT route here.
Top up from $5. No subscription. No token markup.