How Much Does It Cost to Render 1,000+ Frames on a GPU Render Farm?
The cost of a long animation is frame count × average frame time × hourly price. Worked through with iRender's real prices and recorded tests:
- Recorded tests: 360 Redshift frames took 1 h 18 min on 8x RTX 4090, billed $70.20; 300 Octane frames took 1 h 18 min on 4x RTX 5090 ($49.40).
- Long sessions are cheaper: Sessions of 3 hours or more get 10% off the whole session; multi-GPU plans are billed per minute.
- Estimate before you commit: Render 2 or 3 representative frames, average the time, then multiply out. Heavy frames (volumes, motion blur) change the result most.
- SaaS or IaaS: Distributed SaaS farms spread frames over many nodes for short turnaround; IaaS servers suit scenes you need to open and test first.
- First top-up: A 100% bonus on the first top-up (minimum $50) doubles the credit available for the first project.
For a long animation, the hourly price alone tells you little. What matters is the total: how long each frame takes on the server, times the number of frames, times the price. This guide works that out with iRender’s current prices and two recorded benchmark runs, and shows how to estimate your own project before you start.
Prices checked against the iRender price page, November 2026.
1. How do you calculate the cost of a long animation?
Cost = number of frames × average frame time × hourly price. Average frame time is the only part you have to measure, and it depends on your scene, render engine and server.
- Pick 2 or 3 representative frames, including the heaviest shot.
- Render them on the server you plan to use and note the times.
- Average the frame time, multiply by the frame count to get hours.
- Multiply by the hourly price, and apply the 10% discount if the session runs 3 hours or more.
Checklist (copy or follow line by line):
Cost = frames × average seconds per frame ÷ 3,600 × hourly price
−10% when the session runs 3 hours or more
Example (recorded 9S test): 360 frames, 1 h 21 min billed × $52/h = $70.20
→ ≈ $0.20 per frame
Your job: ______ frames × ______ s ÷ 3,600 = ______ h × $______ = $______
2. What do iRender multi-GPU servers cost per hour?
iRender multi-GPU server prices
From the iRender price page. Billed per minute on 2x, 4x and 8x servers.
| Plan | GPUs | Per hour | From 3 hours (−10%) |
|---|---|---|---|
| 4S | 2x RTX 4090 (24 GB each) | $15.00 | $13.50 |
| 5S | 4x RTX 4090 | $30.00 | $27.00 |
| 9S | 8x RTX 4090 | $52.00 | $46.80 |
| 4i | 2x RTX 5090 (32 GB each) | $20.00 | $18.00 |
| 5i | 4x RTX 5090 | $38.00 | $34.20 |
| 9i | 8x RTX 5090 | $62.00 | $55.80 |
Sources: iRender price page.
3. What did 300 and 360 frames actually cost in recorded tests?
Two of our recorded benchmark videos give real frame counts and times. The costs below are simply time multiplied by the hourly price.
Cost per frame from recorded iRender tests
Times from iRender benchmark videos; costs = time × hourly price. Scaling to 1,500 frames is an illustrative example that assumes the same average frame time.
| Test | Measured | Cost of the test | Illustrative: 1,500 frames |
|---|---|---|---|
| C4D + Redshift, 360 frames 8x RTX 4090 (9S) | 1 h 18 min 13 s rendered, 1 h 21 min billed | Billed $70.20 (≈ $0.20 per frame) | ≈ 5.6 h → $292.50, minus 10% = ≈ $263.25 |
| C4D + Octane, 300 frames 4x RTX 5090 (5i) | 1 h 18 min | 1.3 h × $38 = $49.40 (≈ $0.16 per frame) | 6.5 h → $247.00, minus 10% = $222.30 |
The two recorded runs used in the worked examples.
More Redshift render times on 2x, 4x and 8x servers are in our Redshift benchmark.
4. What makes a 1,000-frame render cost more than expected?
- Uneven shots: a few heavy shots with volumes, motion blur or deep reflections can double the average frame time. Always include them in the test.
- VRAM limits: VRAM is not pooled across GPUs, so every GPU must hold the scene. A scene that spills out of GPU memory slows down sharply.
- Re-renders: fixing a wrong camera, missing texture or color setting after the full run doubles the cost. Check a few finished frames before rendering the whole sequence.
- Idle time: on a rented server you pay while it is on, so upload files to storage before starting the server and shut it down when rendering ends.
A note on sources about VRAM: Maxon’s support article states that Redshift does not combine the VRAM when using multiple GPUs. Maxon’s 3D rendering product page also lists “multi-GPU memory pooling” and says Redshift shares out-of-core memory across multiple cards. This article follows the support article: plan for every GPU holding the scene in its own VRAM. Check Maxon’s latest documentation for your Redshift version.
5. Should you use a SaaS or an IaaS render farm for 1,000+ frames?
A SaaS render farm distributes frames across many machines at once, which usually gives the shortest turnaround for a finished, packaged scene, and you pay per job. An IaaS render farm such as iRender rents you a whole multi-GPU server by remote desktop. Frames render in sequence on that machine, but you can open the scene, install the exact plugins, test frames and fix problems before the full run. Choose SaaS when the scene is final and turnaround matters most; choose IaaS when the scene still needs testing or uses plugins and versions a SaaS farm does not support.
On iRender, Cinema 4D, Redshift and Octane are pre-installed on every server, and project files can be uploaded to iRender Drive free of charge before you start the server. See the Cinema 4D Redshift render farm and Octane render farm pages for plans.
6. Render farm cost FAQ
How much does it cost to render 1,000 frames on a GPU render farm?
It depends on the average frame time. As a worked example from a recorded iRender test, Redshift rendered 360 frames on 8x RTX 4090 at $52/hour with 1 h 21 min billed ($70.20), about $0.20 per frame, so 1,000 similar frames would be about $195, or about $175.50 after the 10% discount for sessions of 3 hours or more.
How do I estimate render farm cost for my animation?
Render 2 or 3 representative frames on the server you plan to use, average the frame time, multiply by the number of frames to get hours, then multiply by the hourly price. Include heavy shots, since they usually dominate the total.
Is 8x GPU always cheaper per frame than 4x GPU?
Not always. An 8x server costs more per hour, so it is cheaper per frame only if the engine scales well enough on your scene. Test the same frame on both and compare cost per frame, not just speed.
Does iRender bill per minute?
Yes for multi-GPU plans (4S, 5S, 9S, 4i, 5i, 9i), which are billed per minute from the start. The single-GPU plans 3S and 3i have a one-hour minimum, then per minute.
Are there discounts for long renders on iRender?
Sessions of 3 hours or more get 10% off the whole session, including the first 3 hours. Daily, weekly and monthly rentals use the same 10% discount.
Is a distributed SaaS farm faster for 1,000+ frames?
Often, because frames are spread over many machines at once, which shortens turnaround. An IaaS server renders frames in sequence on one multi-GPU machine but lets you open the scene, test and fix it first. Compare both on turnaround, total cost and how much testing the scene still needs.
