What is your AI inference actually costing?
Most real products do not buy from one vendor. They run a frontier text model, an image model, something for video or voice, and a cheap model for the boring half of the traffic — across three or four bills that never get compared. Build the whole basket here and see it in one number.
Build your basket
What you need
picks the planYour model spendoptionaldoes not change the planWhat your providers bill you today. Add it to see the subscription beside it — the two are separate bills and are never added together.
- $2 in / $10 out
- $0.03
- $0.084
- $80
The plan that fits
INFRO Startup
$79
a month — smallest plan that covers 250,000 requests a month
Charged monthly
$79
every month, or $780 once a year
Chosen by requests, people, projects, retention and the capabilities you ticked — the same entitlements the gateway enforces. Your monthly selection travels with the button above.
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.
- Media usually dominates before anyone notices
- Text spend is visible because it is measured in familiar units. Video measured in seconds and speech measured in characters tend to overtake it quietly — a single feature that renders clips can outgrow an entire chat product's token bill in a quarter.
- The cheap half of your traffic is worth finding
- In most products a large share of calls are classification, extraction, routing, or moderation running on a frontier model because that was the default. Those are the calls where a tier change costs nothing in quality and a lot in spend.
- An estimate is a floor, not a quote
- Cached input, prompt length distribution, image sizes, retries, and failed generations all move a real bill. This tool prices a basket at published rates so you can see the shape; a price estimate reads your actual usage and prices the real thing.
The models in this calculator
The vendor's own published rate, per unit. This is what they will bill your account — INFRO adds nothing to it, and uses these rates to estimate your spend unless you enter your own negotiated ones.
| Model | Task | Vendor rate | INFRO adds |
|---|---|---|---|
| Claude Fable 5anthropic/claude-fable-5 | Frontier reasoning & agents | $20 /1M tok | $0 |
| Claude Fable 5.1anthropic/claude-fable-5.1 | Frontier reasoning & agents | $20 /1M tok | $0 |
| GPT-6 Astraopenai/gpt-6-astra | Frontier reasoning | $20 /1M tok | $0 |
| GPT-6 Solopenai/gpt-6-sol | Everyday reasoning | $4.00 /1M tok | $0 |
| GPT-6 Lunaopenai/gpt-6-luna | High-volume execution | $0.20 /1M tok | $0 |
| GPT-5.6 Solopenai/gpt-5.6-sol | Frontier reasoning | $8.00 /1M tok | $0 |
| Gemini 3.1 Progoogle/gemini-3.1-pro | Multimodal reasoning | $4.50 /1M tok | $0 |
| Claude Opus 5.5anthropic/claude-opus-5.5 | Coding & complex work | $8.00 /1M tok | $0 |
| Claude Opus 5anthropic/claude-opus-5 | Coding & complex work | $10 /1M tok | $0 |
| Grok 4.7xai/grok-4.7 | Agents & knowledge work | $3.00 /1M tok | $0 |
| Grok 4.6xai/grok-4.6 | Agents & knowledge work | $3.00 /1M tok | $0 |
| Claude Sonnet 5.5anthropic/claude-sonnet-5.5 | Production workhorse | $4.00 /1M tok | $0 |
| Claude Sonnet 5anthropic/claude-sonnet-5 | Production workhorse | $4.00 /1M tok | $0 |
| GPT-5.6 Terraopenai/gpt-5.6-terra | Everyday reasoning | $4.50 /1M tok | $0 |
| Gemini 3.8 Flashgoogle/gemini-3.8-flash | Fast agents & coding | $1.50 /1M tok | $0 |
| Gemini 3.7 Flashgoogle/gemini-3.7-flash | Fast agents & coding | $1.50 /1M tok | $0 |
| GPT-5.6 Lunaopenai/gpt-5.6-luna | High-volume execution | $0.45 /1M tok | $0 |
| Claude Haiku 4.5anthropic/claude-haiku-4.5 | Fast, low-cost chat | $2.00 /1M tok | $0 |
| gpt-oss-120bopenai/gpt-oss-120b | Open-weight reasoning | $0.263 /1M tok | $0 |
| DeepSeek Flash (V4.1)deepseek/deepseek-flash | Budget reasoning | $0.525 /1M tok | $0 |
| DeepSeek V4 Prodeepseek/deepseek-v4-pro | Open-weight reasoning | $1.98 /1M tok | $0 |
| DeepSeek V4 Flashdeepseek/deepseek-v4-flash | Budget reasoning | $0.525 /1M tok | $0 |
| Kimi K3moonshot/kimi-k3 | Long-context agents | $6.00 /1M tok | $0 |
| Kimi K2.7 Codemoonshot/kimi-k2.7-code | Code generation | $1.71 /1M tok | $0 |
Showing 24 of 287.
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.