FX Balance and Cost Analytics

FOXI Studio uses one shared FX balance for paid operations: text generation, voiceover, images, AI video, cloud rendering, WordPress articles, and other features. Users do not need separate balances across many AI services: FOXI routes operations through internal models and shows total cost in one interface.

1 FX is approximately $0.01. So 100 FX is about $1, and 200 FX is about $2. This makes it easier to plan the cost of one video or a whole series.

Where to open costs

The balance is visible in the top panel of FOXI Studio. Click the balance block or open Tokens and costs — this window shows available funds, hold, operation history, filters, and pricing.

In this window you can understand:

• how many FX are currently available;
• how many FX are temporarily held for active jobs;
• how much was spent by project;
• which tasks are the most expensive;
• which stages create most costs;
• which models are used;
• where retries and unnecessary generations happened.

Available, spent, and hold

FOXI has three important money states:

State	Meaning
Available	FX that can be used for new tasks.
Spent	FX already charged for completed operations.
Hold	Temporary reserve for an active job. After completion, the unused part is returned.

Hold is needed so a job does not stop halfway because of insufficient funds. For example, if FOXI starts several video clips or a render, it may temporarily reserve more FX than will actually be charged.

Why video cost can differ

Videos with the same duration can cost different amounts. Price depends not only on length, but also on enabled features.

Main factors:

• script length;
• selected text model;
• voiceover character count;
• voice provider;
• number of images;
• image model and quality;
• number of AI video clips;
• RunWay/VEO mode;
• Builder resolution: 1080p, 2K, or 4K;
• subtitles, avatar, overlays, Anti-Detect;
• WordPress article rewriting;
• retries after errors.

For many regular videos, FOXI can keep production cost around $1–2 if unnecessary expensive options are not enabled. But a video with 4K, many AI clips, avatar, and multiple regenerations will cost more.

Cost table

In the cost table, each operation is recorded separately: project, task/video, stage, service, quantity, cost, and time.

Read the table like this:

• Project — which channel/project the operation belongs to.
• Task / video — which topic or video was processed.
• Operation / stage — text, voice, images, video, Builder, WordPress.
• Service / model — which AI model or cloud service was used.
• Quantity — tokens, characters, images, clips, minutes, or operations.
• Cost — how many FX were charged.
• Retry — whether the operation was a retry after an error.

If a video became expensive, first find the most expensive stage. Do not guess — read the table.

Filters: project, task, stage, period

Filters help avoid mixing all charges together.

Useful scenarios:

• select one project and understand its average cost;
• select one task and see why a specific video became expensive;
• select Video to see AI clip costs;
• select Builder to check render cost;
• view costs for today, week, or month.

For multiple channels, filters are especially important: one project may be cheap and stable, while another is expensive because of 4K or many video clips.

Model prices

The cost window may show a generation pricing table. It gives FX reference prices for text, voiceover, images, video, rendering, and other operations.

Important: the real video cost is the sum of all operations. Even if one text request is cheap, the project may become expensive because of many images, clips, or retries.

Comparison with official rates

FOXI may show a comparison with official model rates when such data is available. This helps understand where FOXI saves through internal routing, package conditions, or optimized processing.

But rate comparison is not the only criterion. The final channel economics matter more: video cost, time saved, how many channels can run in parallel, and how stable the result is.

How to keep videos around $1–2

Practical starting setup:

1. Render the first video in 1080p.
2. Do not enable 4K before validating the niche.
3. In Step 4, animate 3–5 best images, not every frame.
4. For text-to-video tests, generate 1 clip first.
5. Do not regenerate the whole video because of one bad image.
6. Use Prompt History to improve prompts instead of spending FX on random attempts.
7. Check costs after every test video.

If a video became too expensive, ask the AI Assistant: “Look at the project costs and suggest how to reduce cost without losing too much quality.”

What to do with frequent retries

Retries can be normal during temporary model failures, but many retries indicate a problem.

Check:

• whether delays between requests are too short;
• whether too many generations run in parallel;
• whether the external service has enough quota;
• whether keys and OAuth are correct;
• whether the prompt is too complex;
• whether the selected provider is overloaded.

Retries may also spend FX because some operations were already sent to a model. If you see many retries, fix settings instead of continuing a large series.

Multiple projects and channels

FOXI can run several projects in parallel. Cost analytics helps identify which channels are economically efficient.

Compare:

• average video cost per project;
• niche test cost;
• how many FX are lost on errors;
• which formats require expensive video clips;
• where Builder or 4K creates extra load;
• which projects can be scaled.

If one channel consistently creates videos cheaper and faster, it is easier to scale.

Topping up balance

You can top up FX in the balance window. Minimum amount and available payment methods depend on the current account and plan.

Tip: before a large series, first estimate the cost of one test video, then multiply it by the number of videos and add a reserve for retries and regenerations.

Common issues

Balance exists, but the task does not start

Check hold: some funds may be reserved for another active task. Also check plan and feature limits.

Funds are held longer than expected

Check the active task log. If the task ended with an error, refresh project state or ask the AI Assistant with the log text.

A video costs more than expected

Open the task filter and find the most expensive stage. Usually the reason is AI video, 4K Builder, avatar, Anti-Detect, many images, or retries.

It is unclear why FX were charged

Check operation details: stage, model, quantity, and time. If a row is unclear, send it to the AI Assistant.

Need to estimate a series budget

Create 1–3 test videos, check the average cost, and multiply by the planned video count. Do not plan a series only from theoretical model prices.

Result

Cost analytics is not just a charge table; it is a production management tool. It shows where FOXI spends FX, which projects are profitable, which settings make videos more expensive, and what can be optimized before scaling a channel.