Vision, audio, agents: what each capability adds to the bill

A FinOps guide by OptimNow · prices from the live OptimToken catalogue

A product team says the feature needs to read screenshots. Someone then asks what vision costs, and the catalogue appears to answer: models that accept images carry a median blended price 2.88x the models that do not. That number is real and it is the wrong one to budget against. It compares vendors, not capabilities. This guide separates the two on 285 priced models. The cost that remains is the size of the shortlist each requirement leaves.

Three capabilities worth measuring, one worth skipping

OptimToken reads capability tags from upstream model metadata. Image input and audio come from the model's declared input and output modalities. Tool use comes from the parameters the model accepts. Those three are facts about the model, so they can be priced.

The Code tag cannot, and this guide leaves it out. OptimToken assigns it partly from a price threshold, so asking what Code adds to the bill would return the threshold. Quoting a code premium from this catalogue would be circular. Saying so beats dropping it quietly.

The catalogue-wide premium is a mix effect

Capabilities cluster with vendors. A house whose entire line-up is multimodal and priced for enterprises puts every one of its models on the with-vision side. A house selling small text models puts every one of its models on the other. The gap between the two sides is the gap between those two houses.

Compare inside one vendor's catalogue instead and the premium shrinks. Of the providers listing at least 2 models on each side, the median ratio between their vision models and their text-only models is 1.67x, against 2.88x across the catalogue. In 2 of 8 the ratio sits below 1, meaning the vendor's vision models are the cheaper half of its own line-up.

The spread across vendors is wide, from 0.57x to 2.23x, so image input is not free everywhere. The vendor you buy from decides more of the bill than the capability does.

Counter-example: dropping a capability can raise the bill

The usual FinOps instinct is to treat model selection as rightsizing, and to stop paying for what the feature does not use. On this catalogue that instinct misfires at 2 of the 8 vendors with enough models to compare. At Zhipu the median vision model costs $0.986 per 1M blended tokens against $1.720 for its text-only models, so removing image input from the requirement moves the shortlist to the more expensive half.

At Google, the cheapest model accepting images is Gemma 3 4B at $0.085 per 1M blended tokens. Drop the image requirement and the cheapest thing on the same shelf is Gemma 2 27B at $0.650, a 7.65x gap in the wrong direction. Compare them side by side. The vendor set the price by the model's position in the line-up. The capability came with it.

What a requirement costs instead: the shortlist

If the rate card does not price capability, something else does. Every requirement removes models from consideration. A thinner shortlist costs you the competition that was holding the price down.

Read as a shortlist, the four requirements are not comparable at all. Tool use leaves 253 of 285 models, 89% of the catalogue. The cheapest model carrying it is Mistral Nemo at $0.027 per 1M blended tokens, the cheapest model in the catalogue outright. Requiring tool use costs nothing at the floor. The budget option already has it.

Worked example: the audio requirement

Audio behaves differently. It is the requirement worth planning around. Across the catalogue, 24 of 285 models accept or produce it, 8% of the total. The catalogue-wide price ratio is only 1.54x, so on a rate-card reading audio looks close to free. The cheapest model that handles audio is MiMo-V2.6-Flash at $0.238 per 1M blended tokens, against $0.027 for Mistral Nemo with no capability requirement at all. Adding audio to the specification multiplies the floor by 8.91x.

Concentration compounds it. Google publishes 13 of the 24 audio models, 54% of the shortlist, spread across 7 providers in total. A requirement met by one vendor on most of the shortlist is a negotiating position as much as a technical choice, and it belongs in the risk register next to the cost estimate. Where audio is a minority of requests, a router is the cheaper structure. Send the audio requests to a model that handles them and leave the rest on the catalogue floor. Paying the audio floor on every call buys audio for calls that carry none.

How to use this when choosing a model

These steps assume a team that already tags requests by type, the Walk stage in FinOps maturity terms. A team still reading one monthly invoice has nothing to route yet, and request-level tagging comes first.

• Specify capabilities per request type rather than per feature. One multimodal model for a workload that is 5% images sets the floor for the other 95%.
• Check the vendor's own ladder before paying for a capability. Within a line-up the premium ran a median 1.67x for image input, and went the other way at 2 of 8 vendors.
• Count the shortlist, then look at its concentration. A requirement that leaves 24 models, most from one vendor, is a different risk than one that leaves 253.
• Set the quality bar before the capability list. The bar belongs to the product owner; the model choice is the cheapest path that clears it. Capability tags say what a model accepts. They say nothing about how well it does the job.

These are published list prices for the model's own API, on a 30/70 input-output blend (the formula is here). They do not include what a capability costs in tokens once it is used. An image is billed as input tokens when it reaches the model, and a long screenshot is a long prompt. Price the payload with that formula; this guide prices the shortlist only.

Figures on this page come from the 2026-10-04 price snapshot and rebuild nightly.

The FinOps reasoning behind this guide comes from OptimNow's open-source practice library: github.com/OptimNow/cloud-finops-skills. If you want your own model shortlist sized against the capabilities your workload uses, OptimNow runs that review.

Data snapshot: 2026-10-04 · Compare all models · Cloud compute pricing · All guides