What are you spending on OpenAI?
OpenAI bills differently by modality: text per token with separate input and output rates, images per image by size, video per second, and speech per character. That is four units on one invoice, which is why a single monthly figure so rarely tells you where the money went. Split your spend by model below.
Build your basket
Your monthly AI spend
3 models- −35%
- −40%
- −50%
Enter what you spend today at each provider's direct price. Savings are computed per model from the rates published on this page — models we price the same as the provider show no saving, and are left in rather than quietly dropped to flatter the total.
Your estimate
Current spend
$4,600
per month
Estimated with INFRO
$2,870
per month
Potential savings
$1,730
38% blended across your basket
Over twelve months
$20,760
same basket, same rates
38% of current spend
An estimate, not a quote. It prices your basket at the reference and INFRO rates published on this page — your real mix of prompt lengths, cached tokens, and image sizes will move it.
What moves this bill more than the vendor does.
A calculator prices a basket. These are the things that make a real invoice differ from one — worth knowing before you treat any estimate, including ours, as a number.
- Input and output are not the same price
- Output tokens cost several times what input tokens cost across the GPT-5.6 family, so a prompt-heavy retrieval workload and a generation-heavy one with the same token count produce very different bills. The blended figure here assumes roughly 3:1 input to output — if you run long documents through short answers, your real rate is lower than the blend.
- Cached input changes the arithmetic
- Repeated prompt prefixes are billed at a reduced rate once cached. An agent that resends a large system prompt on every turn is the workload where this matters most, and it is the single biggest reason a measured bill differs from an estimate like this one.
- Sora is billed per second, not per render
- Video cost scales with duration and resolution rather than with the number of jobs, so a batch of short clips and one long render at the same total seconds cost roughly the same. Estimate video by seconds per month, not by clips.
Every OpenAI model we carry
Direct price beside ours, per unit. Rows priced at parity are left in rather than dropped.
| Model | Task | Direct | INFRO | Save |
|---|---|---|---|---|
| GPT-5.6 Solopenai/gpt-5.6-sol | Frontier reasoning | $11 | $7.88 /1M tok | −30% |
| GPT-5.6 Terraopenai/gpt-5.6-terra | Everyday reasoning | $4.50 | $2.92 /1M tok | −35% |
| GPT-5.6 Lunaopenai/gpt-5.6-luna | High-volume execution | $0.45 | $0.247 /1M tok | −45% |
| GPT Image 2openai/gpt-image-2 | Image generation & editing | $0.06 | $0.036 /image | −40% |
| Sora 2openai/sora-2 | Video generation with audio | $0.10 | $0.06 /second | −40% |
| Whisper v3 Turboopenai/whisper-v3-turbo | Transcription | $0.36 | $0.18 /audio hour | −50% |
| TTS-2openai/tts-2 | Text to speech | $30 | $26 /1M chars | −13% |
USD, illustrative launch rates reconciled on 2026-08-24. Live rates come from GET /v1/models once your key is active.
An estimate is a floor. Want the real number?
Send a usage export and we will price your actual traffic — with the arithmetic shown, and the caveats named.