AI and automation

AI Image and Video Generation Cost Calculator

Compare monthly visual AI spend across subscription plans, credits, GPU-hour overage, and pay-as-you-go image or video rates.

When to use this

Use it before a campaign turns into a credit burn

This calculator is for creators, marketing teams, product teams, and founders planning AI images, short videos, thumbnails, ads, mockups, or batch visual experiments.

Default result

The server-rendered creator baseline shows Google Vertex AI as cheapest at $3.60 per month.

Compare a visual workload

Start with a preset, then replace the output counts, video seconds, GPU hours, credit rates, and provider settings.

Images per month Videos per month Avg video seconds

Midjourney Plan Billing mode Planned fast GPU hours

Runway Plan Video credits per second Manual overage cost per credit ($)

Google Vertex AI Imagen rate Veo rate

Manual API modes Include OpenAI GPT-Image-2 manual estimate

Input tokens/image Output tokens/image Cached input share (%) Input price / 1M ($) Cached input / 1M ($) Output price / 1M ($)

Include manual Sora plan estimate

Sora plan cost ($) Included minutes Overage per minute ($)

Sora stays manual until current official plan limits and pricing are verified. No Sora price is hardcoded here.

Output volume

82 assets

40 generated video seconds.

Midjourney overage

$32.00

Extra fast GPU hours are charged at the sourced $4/hr rate.

Runway credits

1,120.6 used

0 overage credits in the default scenario.

Worked example: creator baseline

The default scenario uses 80 images, 2 videos, and 20 seconds per video. That is 82 total assets and 40 generated video seconds.

Midjourney Pro monthly is $60 and includes 30 fast GPU hours. The scenario plans 38 fast hours, so overage is 38 - 30 = 8 hours. At $4/hr, overage is $32 and total Midjourney cost is $92.00, or $92.00 / 82 = $1.12 per asset.

Runway Pro is $28 per month in the captured annual-billed pricing page. Image credits are 80 x (2,250 / 281) = 640.57. Video credits are 2 x 20 x 12 = 480.00. Total credits are 1,120.57, below the 2,250 included credits, so Runway stays at $28.00, or $0.34 per asset.

Google Vertex AI uses Imagen 4 Fast at $0.02/image and Veo 3.1 Lite 1080p video-only at $0.05/sec. Image cost is 80 x $0.02 = $1.60. Video cost is 40 x $0.05 = $2.00. Total Google cost is $3.60, or $0.04 per asset.

For this exact cost-only setup, Google Vertex AI is the cheapest path. The result does not score output quality, style fit, moderation policy, editing UX, rights, or support burden.

How to estimate AI image and video generation costs for one month

Start from the number of final images, the number of generated videos, and the average video length. Then compare subscription plans, credits, GPU-hour overage, and pay-as-you-go rates on the same monthly basis.

How Midjourney fast GPU hours turn into overage charges

Midjourney cost is plan price plus fast GPU overage. The calculator does not invent a per-image conversion because the cited source gives included fast GPU hours and the $4/hr extra GPU rate, not a direct asset conversion.

How Runway credits are spent across images and seconds

Runway cost depends on included monthly credits and the credit burn of the selected tools. The default uses the sourced image equivalents and a 12 credits/second video rate for Gen-4.5 style video. Override it when your actual Runway tool uses a different rate.

How Google Vertex AI prices Imagen and Veo by resolution and second

Google Vertex AI is modeled as pay-as-you-go: images cost the selected Imagen rate per image, and videos cost the selected Veo rate per generated second. Resolution and audio change the Veo rate.

When OpenAI GPT-Image-2 token costs become useful in visual budgeting

OpenAI GPT-Image-2 is useful when you have measured image input and output tokens for your exact settings. The calculator keeps that path manual because exact per-image token accounting can vary by prompt, image settings, and source updates.

Why a Sora manual mode is required without current official per-video prices

The current source set does not include a verified per-video Sora price table. The Sora block is manual so teams can compare a verified plan only after checking current official app limits and pricing.

How to read cheapest path, tie cases, and overage risk

The cheapest result is a cost screen, not a quality score. If two paths are within 10%, choose by output quality, editability, commercial terms, speed, moderation policy, watermarks, and workflow fit.

What to verify before locking a production budget

Run one dry week, export the actual provider usage, and replace every editable assumption. Check plan changes, private contract terms, regional pricing, rights, review time, and failed-generation rates before committing budget.

Reference data used by the defaults

Topic	Reference point	Source	Date	Note
OpenAI GPT-Image-2 API pricing	Image input $8.00 per 1M tokens, cached image input $2.00 per 1M tokens, image output $30.00 per 1M tokens; text input $5.00 per 1M tokens and cached text input $1.25 per 1M tokens.	OpenAI API Pricing	As of June 19, 2026	Exact per-image token accounting depends on settings. Use manual token estimates and verify on the source before budgeting.
Midjourney subscription and overage	Basic $10, Standard $30, Pro $60, Mega $120 monthly; annual equivalents $8, $24, $48, $96 per month; fast GPU time 3.3, 15, 30, 60 hours; extra GPU time $4/hr.	Midjourney Comparing Plans	As of June 19, 2026	Midjourney does not publish a simple per-image or per-video conversion in the cited source. Enter measured fast GPU hours.
Runway plan credits	Free 125 one-time credits; Standard 625 monthly credits; Pro 2,250; Max 9,500. Pricing page gives seconds/images equivalents by plan.	Runway Pricing	As of June 19, 2026	Plan amounts shown are annual-billed monthly equivalents in the captured pricing page. Verify current billing mode and quotas.
Runway credit rules	Runway credits are used for images, videos, and audio. Gen-4.5 uses 12 credits per second in the help example; web app credits and API credits are separate.	Runway Credits Help	As of June 19, 2026	Credit rates can vary by tool and model. Use the editable credit assumptions for your actual workflow.
Google Imagen and Veo pricing	Imagen 4 Fast $0.02/image, Imagen 4 $0.04/image, Imagen 4 Ultra $0.06/image. Veo 3.1 Lite video-only is $0.03/sec at 720p and $0.05/sec at 1080p; higher tiers cost more.	Google Cloud Agent Platform Pricing	As of June 19, 2026	Prices are listed in USD and can vary by region, contract, or SKU. Verify before budgeting.

Verify each provider page before using the result for production budgeting. The manual OpenAI and Sora fields are intentionally excluded from the default ranking until you enter verified values.

FAQ

Why does this tool include Midjourney fast GPU hours as input?

The cited Midjourney source provides plan fee, included fast GPU time, and overage rate, not a direct per-image or per-video conversion. Enter measured fast hours from your own workflow.

Does Runway credit math depend on resolution and model choice?

Yes. Credits per image and credits per video second can change by model, plan, tool, and quality setting. This calculator starts from the cited plan equivalents and lets you override the video credit rate.

Why can one provider be cheapest by dollars but not by quality?

This tool computes cost only. Rendering quality, prompt control, visual consistency, rights, moderation policy, watermarks, speed, and editing workflow must be evaluated separately.

Why is Sora not fully auto-calculated?

The current source set does not provide a verified per-video Sora price or API-style rate table. Sora stays as a manual mode until current official plan limits and pricing are verified.

Can I compare subscriptions with pay-as-you-go pricing?

Yes. Midjourney and Runway are subscription-and-quota style in this calculator, while Google Vertex AI is usage-based. The result ranks monthly planning cost across both patterns.

Can this be used for real budget approvals?

Use it as a planning model, then add storage, review time, failed generations, legal review, procurement rules, and engineering or editing overhead before approving a production budget.

Decision path

What to do next

API budget