Rendi Pricing (2026): Pick by Bytes, Not Video Length

Rendi pricing starts at $0, but the first paid plan is $25 per month for 250 GB of input-plus-output processing, 50 GB-month of storage, 4 vCPUs, and a 10-minute command cap. Most pipelines should buy the smallest tier that clears four independent constraints: processed bytes, retained storage, longest command runtime, and concurrent CPU demand, not video duration.

Rendi Pricing at a Glance

Rendi's $25 plan is the sensible default for a working automation, provided every command finishes within 10 minutes. The Free plan is unusually useful, but its 1-minute command cap will force many production jobs onto Pro before the 50 GB processing allowance runs out.

I verified every row below against Rendi's live pricing page on 11 September 2026.

All displayed prices are monthly. Rendi publishes no annual billing option or annual discount on this page, so there is no annual commitment to normalize against the monthly figures.

Every paid configuration adds unlimited commands per minute, priority processing, private file storage with presigned URLs, dynamic input and output files, chained commands, API webhooks, an expert consultation session, and email support. Enterprise adds uptime SLAs, premium support, SOC 2, custom DPA and SLA terms, custom API limits, and dedicated infrastructure.

The ladder is linear rather than discount-led. The $25 plan includes 250 GB, and Rendi describes Pro pricing as low as $0.10 per processed GB. Moving to a larger row mainly buys more processing, storage, and account-wide vCPU capacity. It does not create an obvious volume discount in the published table.

Rendi FFmpeg API Cost: Input Plus Output

Rendi bills the bytes that cross both sides of a command. The formula is simple: processed GB = input GB + output GB.

Rendi's own example sends a 1 GB input into FFmpeg and produces a 0.5 GB output. That job consumes 1.5 GB of processing quota. If one command creates several outputs, the useful budgeting habit is to count the input and every output before choosing a plan.

This is why video duration alone cannot predict the bill. A compact source with a compact output may be cheap even when the timeline is long. A high-bitrate source with several renditions can consume far more quota despite a shorter timeline. Duration still matters for the 1-minute, 10-minute, or unlimited command-runtime gate, but it is not the pricing unit.

For the vendor's 1.5 GB example, the outcome math looks like this:

• Free covers 33 complete jobs because 50 divided by 1.5 is 33.3.
• The $25 plan covers 166 complete jobs because 250 divided by 1.5 is 166.7.
• At full use, each completed example job consumes $0.15 of the subscription price.
• At 100 jobs, the workload consumes 150 GB, so the same $25 plan costs an effective $0.25 per completed job.

That last line matters. A $0.10-per-GB headline is the capacity rate, not necessarily your realized rate. Paying $25 and using only a small fraction of the allowance makes the effective rate much higher. Rendi rewards a predictable monthly workload, not occasional use that happens to fit the technical limits.

FFmpeg Pricing Is Hosting, Not a Software License

FFmpeg is free software. Rendi charges for managed compute, storage, API orchestration, scaling, and delivery around it. You are paying to submit an FFmpeg command over HTTP and receive stored outputs without maintaining video workers yourself.

Rendi also states that it charges no separate ingress or egress fee and applies no special billing unit for encoding choice, resolution, or video duration. Those omissions simplify forecasting, but they do not make processing or storage unlimited.

Who Can Stay on Rendi Free?

Stay on Free indefinitely when every workload fits all five operational limits, not merely the $0 price: 50 GB of input-plus-output processing per month, 5 GB-month of storage, 4 account vCPUs, a 1-minute maximum per command, and 4 command submissions per minute.

That profile fits a prototype, an internal utility, or a low-frequency automation that trims, probes, extracts, or converts modest files quickly. Using Rendi's 1.5 GB example, 33 completed jobs fit the processing allowance. The plan also includes file storage and analysis, multiple API environments, Zapier and Make integrations, and community support.

The 1-minute gate is the usual wall. It measures how long the command runs on Rendi's compute, not how long the source video plays. A long source can pass if the operation is fast, while a short source can fail if the filter graph or codec work is heavy. Measure actual command runtime before assuming a clip's duration qualifies it for Free.

Free requires a card and a $5 verification charge. Rendi says the charge is refunded after the card is validated and that the Free tier has no recurring fee. It is therefore free after verification, but it is not a no-card sandbox.

How to Choose a Paid Tier

Choose Pro by the first constraint your pipeline exceeds. Processing volume gets the attention, but runtime, retained storage, and concurrent compute can force a larger bill with plenty of byte quota left.

Rendi Pro Plan: Four Constraints Decide the Tier

Processing is the monthly sum of input and output media. Move up the 10-minute ladder at 250 GB, 500 GB, 1 TB, and beyond only when the next published processing allowance is needed.

Runtime is the maximum wall-clock time for one command. If any necessary command cannot finish within 10 minutes, the unlimited-runtime family starts at $100, even if total monthly processing is tiny.

Storage is a separate GB-month allowance. The $25 plan includes 50 GB-month, while the 4-vCPU unlimited-runtime plan includes 100 GB-month. A pipeline that retains outputs can hit storage before processing.

vCPUs govern account-wide concurrency. Rendi lets multiple commands run in parallel while their combined vCPU use stays within the account quota. A single command can use no more than 32 vCPUs on the published plans. The 64-to-256-vCPU rows therefore buy parallel capacity, not one command with 64-to-256 vCPUs.

The most revealing choice sits at $100. Both options include 1 TB of processing:

• The 10-minute option gives 16 vCPUs and 250 GB-month of storage.
• The unlimited-runtime option gives 4 vCPUs and 100 GB-month of storage.

At the same price, the 10-minute configuration provides 4x the account vCPU quota and 2.5x the storage. The unlimited configuration trades both away so one command can keep running. A batch of short parallel jobs belongs on the 16-vCPU row. A serial long encode belongs on the 4-vCPU unlimited row.

Rendi Processing Limits and Hidden Costs

The biggest hidden cost is not egress. It is paying for the wrong constraint because the public page leaves quota-exhaustion behavior unspecified.

Rendi says there are no ingress or egress fees. That removes a common cloud-media surprise, but the live page does not publish a processing overage rate, a storage overage rate, or whether a job stops, queues, or triggers an upgrade when the account reaches a quota. Treat the displayed allowance as a planning boundary and ask Rendi to confirm production behavior before launch.

The page also publishes no rollover promise. Processing is labeled monthly and storage is labeled in GB-month, so unused capacity should not be treated as a banked credit without written confirmation. There is no published annual discount to offset an underused month.

Storage needs active housekeeping. Rendi's API reference says output files remain in its hosted storage indefinitely until you delete them through the deletion endpoints. That is convenient for stable output URLs, but an automation that never deletes old outputs keeps a growing retained footprint against a storage allowance expressed in GB-month.

A safe production design copies durable assets to the storage system you already operate, verifies the copy, then deletes Rendi's output when its working life ends. If Rendi is the permanent origin, estimate average retained data separately from processed bytes. Processing the same file twice and storing two outputs affects both meters in different ways.

The published plans describe AMD CPU capacity. A workflow that depends on GPU-specific encoders should not assume the standard rows support it. Confirm a custom arrangement or choose a service that publishes GPU access.

Refund language is narrow too. The $5 card-verification charge is refundable, and Rendi promises a money-back guarantee for the expert consultation session. The live pricing page does not state a general subscription-refund window. Buy the first paid month as a measured production trial, not as if a broad refund policy were guaranteed.

FFmpeg API Pricing: Rendi vs Two Alternatives

Rendi wins when byte volume is predictable and commands need between 1 and 10 minutes. Very Good FFmpeg is stronger for variable usage and long or GPU-backed jobs. RenderIO is easier to forecast when command count is stable but file size varies.

Very Good FFmpeg

Very Good FFmpeg is another raw FFmpeg API, but it has no monthly subscription minimum. Its first 2 GB are free without a card, then its published rates are $0.50 per GB from 0 to 10 GB, $0.10 from 10 to 100 GB, and $0.08 above 100 GB. It meters input plus output, matching Rendi's underlying unit.

The operational package is different: CPU jobs use 16 vCPUs, Nvidia L4 GPU jobs are available, and a job can run for up to 6 hours. The vendor also says there are no egress or storage fees. That combination is a better fit when monthly volume is spiky, a command can run far beyond 10 minutes, or GPU execution is required.

Its pricing page does not explain whether volume brackets are marginal or retroactive. That ambiguity matters for an exact invoice forecast, so use the published rates for normalization and confirm the bracket calculation before switching.

RenderIO

RenderIO charges by command rather than by byte. Starter is $12 for 500 commands with a 1-minute maximum and 5 GB storage. Growth is $39 for 1,000 commands with a 5-minute maximum and 10 GB storage. Business is $99 for 20,000 commands with a 90-minute maximum, 200 GB storage, and webhooks.

One API call with one FFmpeg command counts as one command regardless of file size. Overage is explicit at $0.08 per command on Starter, $0.05 on Growth, and $0.02 on Business. There is no permanent free tier, but Starter has a 3-day trial and all plans advertise a 30-day money-back guarantee.

Now normalize all three around 100 one-command jobs, each using Rendi's own 1 GB input and 0.5 GB output example. The month represents 150 GB of processed media.

• Rendi needs the $25 plan and lands at $0.25 per job, provided each command finishes within 10 minutes.
• Very Good FFmpeg lands at $12, or $0.12 per job, when normalized at its published $0.08 rate above 100 GB. Treat that as a rate comparison, not an invoice quote, until the vendor confirms how brackets apply.
• RenderIO Starter costs $12, or $0.12 per job, if every command finishes within 1 minute. If jobs need up to 5 minutes, Growth costs $39, or $0.39 per job. If they need up to 90 minutes, Business costs $99, or $0.99 per job.

The cheapest result changes with runtime and meter shape. RenderIO Starter beats Rendi in this example only when the 1-minute ceiling holds. Rendi beats RenderIO Growth when jobs take 1 to 5 minutes. Very Good FFmpeg avoids the subscription floor and has a much longer runtime, but its public bracket wording needs clarification.

Who Should Buy Rendi, and Who Should Skip It?

Buy Rendi when you already know the FFmpeg commands you need and want to run them from a serverless app, Make, Zapier, or another automation without operating video workers. The product is infrastructure, not a visual video editor or a template system.

Skip Rendi when you need a visual timeline, a prebuilt template editor, a published standard GPU tier, a usage-only bill with no monthly floor, or explicit overage behavior. Also separate processing from discovery: if the real problem is finding footage inside a media library, start with these AI video search tools for production teams before buying another processing layer.

Your Monday move is to measure one representative workflow before subscribing.

1. 1

   Measure both byte directions

   Record the total input and output size for each representative command. Multiply by the expected monthly job count.

2. 2

   Record the slowest command

   Use observed wall-clock command time, not source-video duration. Classify the workflow under 1 minute, under 10 minutes, or beyond 10 minutes.

3. 3

   Map concurrency and storage

   Add the vCPUs used by commands that must run simultaneously. Estimate average retained outputs in GB-month and define when files are deleted.

4. 4

   Choose the first row that clears everything

   Select the smallest published plan that passes all four measurements. Ask support how quota exhaustion and storage overage behave before the pipeline handles customer work.

Rendi Pricing FAQ

Is Rendi's FFmpeg API free?

Yes. Free includes 50 GB of processing each month, 5 GB-month of storage, 4 vCPUs, up to 4 commands per minute, and a 1-minute maximum per command. Rendi requires a $5 card-verification charge and says it refunds that charge after successful validation.

Does Rendi offer a student discount?

Rendi's live pricing page does not list a student or education discount. The permanent Free tier is the only publicly documented no-subscription option, so students who exceed it should ask Rendi directly rather than assume special pricing.

What is Rendi's refund policy?

Rendi explicitly refunds the $5 card-verification charge after successful validation and offers a money-back guarantee for its expert consultation session. Its live pricing page does not publish a general refund window for Pro subscriptions.

Has Rendi pricing changed?

The current live table differs from Rendi's May 2026 comparison article. That article says $25 included 400 GB and 16 vCPUs at $0.06 per GB; the live page verified on 11 September 2026 says $25 includes 250 GB and 4 vCPUs, and describes Pro pricing as low as $0.10 per processed GB. Use the live table.

What happens when I exceed Rendi's processing or storage quota?

The live pricing page does not publish an overage rate or say whether jobs stop, queue, or trigger an upgrade. It does state that there are no ingress or egress fees, but that is a separate policy. Confirm quota-exhaustion behavior with Rendi before production traffic arrives.

Does Rendi have annual pricing?

No annual plan or annual discount is listed on the live pricing page. Every public price in the current table is monthly.

Before adding another media infrastructure subscription, get the AI Business Workflow Audit Checklist.