Shotstack vs Very Good FFmpeg: API Docs, Pricing, and Features Compared (2026)

Why compare Shotstack video editing API with a raw FFmpeg REST API?

Shotstack is a cloud video editing API that uses a JSON timeline to describe edits. Very Good FFmpeg is a hosted FFmpeg REST API that runs whatever FFmpeg command you send it.

If you are searching for "Shotstack video editing API FFmpeg REST API docs pricing," you are likely evaluating Shotstack for a video automation project and want to understand how it compares to running raw FFmpeg commands in the cloud. This article compares Shotstack and Very Good FFmpeg across API design, documentation quality, pricing, use cases, and limitations so you can decide which fits your pipeline.

Key takeaways: Shotstack vs Very Good FFmpeg

Shotstack abstracts video editing behind a JSON timeline. Very Good FFmpeg exposes raw FFmpeg directly. The right choice depends on whether you need template-driven video generation or full encoder control.

Here is the short version:

What is Shotstack?

Shotstack is a cloud video editing API that lets you compose videos using JSON. You describe an edit as a timeline with tracks, clips, transitions, titles, and motion effects, send it to the Shotstack API, and it renders the video on cloud infrastructure.

Shotstack was founded in 2019 and is based in Sydney, Australia. It claims 50,000+ developers, 1.1 million videos rendered per month, and customers including Ikea, McDonald's, and Nike. The platform offers three APIs: Edit (render videos/images/audio), Serve (host and manage assets), and Ingest (upload and transform source footage). Each API has a separate base URL and authentication via x-api-key header.

Shotstack openly positions itself as a hosted alternative to FFmpeg. The company has a dedicated page marketing itself as a "cloud hosted alternative to FFmpeg" that acknowledges FFmpeg is powerful but says building web apps with it is challenging due to the steep learning curve and server maintenance requirements.

What is Very Good FFmpeg?

Very Good FFmpeg is a hosted FFmpeg REST API. You send it an FFmpeg command string with input file URLs, and it runs the command on dedicated infrastructure and returns output URLs. No abstraction layer, no JSON DSL, no templates.

Very Good FFmpeg runs every CPU job on a dedicated 16 vCPU / 32 GB DDR5 RAM / NVMe instance with 5+ GHz cores. GPU jobs run on Nvidia RTX and A-series hardware (RTX 4090, RTX 5090, A4000, A5000). There is no shared tenancy or throttling. The maximum single-job runtime is 6 hours, which covers full-length feature films without chunking. Rate limits start at 10 requests per second and increase to 100 with a prepaid balance.

The product supports a full FFmpeg build with standard codecs including H.264, H.265/HEVC, AV1, VP9, ProRes, AAC, and Opus. All standard FFmpeg filters and filter graphs work. Command chaining lets you run multiple FFmpeg commands sequentially in a single request.

How do the two platforms approach video editing differently?

Shotstack abstracts FFmpeg complexity behind a JSON timeline. Very Good FFmpeg exposes FFmpeg directly with no abstraction layer.

With Shotstack, you describe what you want to happen in JSON. You define tracks, place clips on them, add transitions between scenes, layer titles over video, and apply motion effects. Shotstack's rendering engine interprets this timeline and generates the output. You do not write FFmpeg commands. You do not need to understand codecs, pixel formats, or filter graphs.

With Very Good FFmpeg, you write the exact FFmpeg command you want to run. If you know how to use FFmpeg on the command line, you already know how to use Very Good FFmpeg. The API is a thin wrapper that accepts your command, runs it on dedicated hardware, and gives you the output.

This trade-off between ease of use and control is the core difference between the two platforms.

What does Shotstack's API actually look like?

Shotstack exposes a REST API where you POST a JSON body describing a timeline to the Edit endpoint. The JSON defines tracks, clips, assets, transitions, and effects.

Here is a simplified example of what a Shotstack request body looks like:

POST https://api.shotstack.io/edit/v1
x-api-key: YOUR_API_KEY

{
  "timeline": {
    "tracks": [
      {
        "clips": [
          {
            "asset": {
              "type": "video",
              "src": "https://example.com/input1.mp4"
            },
            "start": 0,
            "length": 10
          }
        ]
      }
    ]
  }
}

Shotstack supports many asset types: video, image, text, rich text, audio, shape, luma, caption, HTML, title, and SVG. Merge fields allow data-driven personalization from CRM, CSV, MySQL, or PostgreSQL data sources. Output destinations include Shotstack hosting, Mux, S3, GCS, Google Drive, and Vimeo.

The JSON timeline approach works well for template-style generation where you keep the same structure and swap out assets. For example, generating 1,000 personalized onboarding videos from a CRM dataset with different names, company logos, and background footage.

What does the Very Good FFmpeg API look like?

Very Good FFmpeg uses a single endpoint. You send input file URLs, output file names, and an FFmpeg command string.

POST https://verygoodffmpeg.com/api/ffmpeg
Authorization: Bearer YOUR_API_KEY

{
  "input_files": {
    "input.mp4": "https://your-bucket.s3.amazonaws.com/input.mp4"
  },
  "output_files": ["output.mp4"],
  "ffmpeg_commands": [
    "-i {{input.mp4}} -c:v libx264 -crf 23 {{output.mp4}}"
  ]
}

The API supports an array of FFmpeg commands for chaining. You can transcode to multiple resolutions, extract thumbnails, and package HLS in a single request. The API also supports optional upload endpoints, real-time log streaming in the dashboard, and auto-diagnosis that explains why a command failed.

There is no SDK requirement. Official TypeScript and Python clients are available but optional.

How do the documentation quality and developer experience compare?

Shotstack has polished documentation with a guide, API reference, and SDK support across seven languages. Very Good FFmpeg has concise documentation focused on the FFmpeg command model.

Here is how the docs compare:

Shotstack's docs are more comprehensive if you are new to programmatic video editing. The guide walks you through the timeline concept step by step. The API reference is well-organized across three APIs.

Very Good FFmpeg's docs assume you already know FFmpeg. If you do, there is almost nothing to learn. If you do not, you need to learn FFmpeg CLI separately.

How does the pricing work for each platform?

Shotstack uses per-minute pricing with tiers and credits. Very Good FFmpeg uses per-GB processed pricing with no tiers and no monthly minimum.

Shotstack pricing

Shotstack offers three pricing tiers:

• PAYG: $0.30 per minute of rendered video. Requires a $75 one-time credit pack. Credits roll over up to 3x the monthly allowance. Overage charged at +30%.
• Subscription: $0.20 per minute. Costs $39 per month. Includes the same credit system and overage rules.
• High volume: Custom pricing starting at 50,000 minutes per year.

All plans include 1080p output, a dev sandbox (with watermarks on sandbox renders), white-label SDK, and a maximum render length of 3 hours. Resolution does not affect pricing. A 4K render at 60fps costs the same as 720p at 30fps. One credit equals one minute of video. Videos are rounded down to the second.

Very Good FFmpeg pricing

Very Good FFmpeg charges per GB of data processed (input + output file sizes). The pricing tiers are:

• First 10 GB: per GB (entry rate)
• 10-100 GB: per GB
• 100+ GB: per GB (highest volume discount)

There is no monthly minimum (monthly minimum charge is $0). No subscription fees. No credit packs. No overage penalties. The first 2 GB are free as a trial.

Real cost comparison

Comparing per-minute and per-GB pricing is not straightforward because the units differ. A 60-second 1080p video might be 50-200 MB depending on codec and bitrate. At Shotstack's $0.20/min subscription rate that same video costs $0.20 to render. On Very Good FFmpeg, a 100 MB file at the entry rate would cost significantly less per job.

Very Good FFmpeg is cheaper for raw transcoding workloads. Shotstack's pricing makes more sense when you need the editing features, not just transcoding.

What are the hidden costs and limitations?

Shotstack has several edge cases that increase cost or limit flexibility:

• Overage penalties: If you exceed your credit allowance, overages are charged at +30% on top of the base rate.
• Credit expiration: Unused credits roll over up to 3x the monthly allowance. After that, they expire.
• Sandbox watermarks: All sandbox (stage) renders include a watermark. You must use production to get clean output.
• No free tier: There is no free plan. You pay the $75 credit pack minimum to start testing.
• Max render length: 3 hours per render. Feature-length content requires splitting.
• Limited codec control: You choose from predefined output settings rather than specifying exact encoder flags.

Very Good FFmpeg has fewer hidden costs:

• No overage penalties: Price per GB is fixed per tier with automatic volume discounts.
• No credit expiration: You pay for what you process. No credits to expire.
• Free trial: First 2 GB free with no credit card required.
• Long runtime: 6 hours covers feature-length films.
• Full codec control: Every FFmpeg encoder, pixel format, bitrate, and container is available.

Does Shotstack support FFmpeg filters and custom encoding?

Shotstack has built-in motion effects, transitions, and filters, but it does not expose raw FFmpeg filters. You are limited to the effects Shotstack provides through its JSON timeline model.

Shotstack supports chroma key (green screen), transitions (fade, dissolve, wipe, slide), motion effects (zoom, pan, rotate), and image filters (brightness, contrast, saturation). These are controlled through JSON properties, not FFmpeg filter graph syntax.

If you need a specific FFmpeg filter (like drawtext, lut, curves, eq, hue, colorbalance, frei0r, or custom filter_complex graphs), Shotstack cannot run it. You must use a platform that accepts raw FFmpeg commands.

Very Good FFmpeg supports every FFmpeg filter and encoder. If ffmpeg -filters lists it, the API can run it. This includes hardware-accelerated encoding with h264_nvenc, hevc_nvenc, and GPU-based filters.

What rendering speed does each platform offer?

Shotstack claims fast rendering using parallel split rendering. A 60-second 1080p video renders in approximately 22.8 seconds (about 3x real-time). Shotstack claims this is 7.4x faster than its closest competitor. Rendering time does not increase significantly at higher resolutions because the engine splits work across multiple servers in parallel.

Very Good FFmpeg rendering speed depends on your command and hardware choice. CPU jobs run on dedicated 16-core instances at 5+ GHz. GPU jobs run on Nvidia RTX or A-series hardware. You control threading, preset selection, and hardware acceleration flags. A well-optimized FFmpeg command with the right hardware can match or exceed Shotstack's rendering speed for simple transcoding, but Shotstack's parallel architecture may be faster for complex compositions with multiple layers and effects.

What kind of community presence and support does each platform have?

Shotstack has had limited community traction on Hacker News. Five HN submissions between 2019 and 2022 all received 1-5 points with minimal comments. The Shotstack GitHub organization has 36 repositories and 74 followers. SDK repos have modest star counts: Node SDK has 37 stars, Python SDK has 27, and TypeScript SDK has 1.

Very Good FFmpeg has a smaller GitHub presence but maintains active support channels. The product has been listed on Product Hunt, G2, and several MCP directories. Support is available via email and the dashboard.

Shotstack lists well-known brands (Ikea, McDonald's, Nike, Coca-Cola) as customers on its about page. Very Good FFmpeg features a published case study with Reelo Video.

When should you use Shotstack?

Shotstack works best when your video generation fits the JSON timeline model and you want to avoid FFmpeg complexity.

Good use cases for Shotstack:

• Template-driven personalized video at scale. Generate thousands of videos from CRM data with different fields per customer.
• Teams that include non-developers who understand the timeline metaphor. The white-label editor SDK lets you build a UI on top of the JSON model.
• Projects that need transitions, titles, motion effects, and multi-layer compositions without writing FFmpeg filter graphs.
• Applications where a 3-hour max render length is sufficient and the per-minute cost fits the budget.
• Workflows that benefit from Shotstack's built-in asset hosting and delivery (Serve and Ingest APIs).

When should you use Very Good FFmpeg?

Very Good FFmpeg works best when you need full control over encoding and want simple, predictable pricing.

Good use cases for Very Good FFmpeg:

• Raw transcoding workloads where you need specific codecs, pixel formats, bitrates, or containers.
• Custom FFmpeg filter chains that Shotstack's JSON model cannot express (complex filter graphs, drawtext, frei0r, custom lut).
• Budget-sensitive projects where the higher per-minute cost of Shotstack does not make sense.
• Batch processing at scale where you want auto-applying volume discounts and no credit management.
• Workflows that need longer than 3 hours of continuous processing. Very Good FFmpeg supports up to 6 hours.
• Teams that already know FFmpeg and do not want to learn a new DSL.
• Pipelines that benefit from GPU-accelerated encoding with specific Nvidia encoder flags.

Verdict: Which platform should you pick?

Shotstack is a good product for template-driven video generation where the JSON timeline model matches your workflow and you accept the per-minute pricing. It is especially useful if you need white-label editing capabilities or want to avoid learning FFmpeg entirely.

Very Good FFmpeg is a better fit for raw video processing where you need full FFmpeg control, simpler pricing, and no arbitrary limitations on codec selection or render duration.

The honest answer depends on your pipeline complexity. If your video generation fits a template pattern (same structure, different assets), Shotstack may save development time despite the higher per-minute cost. If you need specific FFmpeg features, custom encoder flags, or complex filter graphs, Very Good FFmpeg is the only option because Shotstack cannot run arbitrary FFmpeg commands.

For pure transcoding workloads, Very Good FFmpeg is cheaper, simpler, and more flexible.

FAQ

Does Shotstack support 4K video?

Yes, at no extra cost. Shotstack charges the same per-minute rate regardless of resolution. A 4K 60fps render costs the same as 720p 30fps.

Does Very Good FFmpeg support 4K?

Yes, with full encoder control. You can use libx264, libx265, prores_ks, VP9, AV1, or any FFmpeg encoder at any resolution.

Can I use Shotstack with any programming language?

Yes. Shotstack provides SDKs for Node, Python, PHP, Ruby, Java, Go, and TypeScript. You can also use the raw REST API with any HTTP client.

Does Very Good FFmpeg have SDKs?

Official TypeScript and Python SDKs are available. An MCP server and Make.com integration are also supported. No SDK is required. The API accepts raw FFmpeg commands over HTTP.

What is the maximum video length for Shotstack?

3 hours per render. High-volume plans may have custom limits.

What is the maximum job runtime for Very Good FFmpeg?

6 hours. This covers feature-length films without chunking.

Does Shotstack support GPU acceleration?

Shotstack does not prominently advertise GPU acceleration. Its rendering speed comes from parallel split rendering across multiple servers.

Does Very Good FFmpeg support GPU acceleration?

Yes. Set the machine type to nvidia to route the job to current-generation Nvidia GPU hardware. You must use GPU-compatible decoders, encoders, and filters in your command (like h264_nvenc, hevc_nvenc) to benefit from GPU acceleration.

Does Shotstack have a free tier?

No free tier. The sandbox (stage) environment allows testing but adds a watermark to all renders. You need to purchase credits (starting at $75) for production output.

Does Very Good FFmpeg have a free tier?

The first 2 GB processed are free. No credit card required to start. After the trial, you pay per GB with no monthly minimum.

Can Shotstack run custom FFmpeg filters?

No. Shotstack has built-in effects (transitions, chroma key, motion, brightness/contrast/saturation) but does not expose raw FFmpeg filters. If you need a specific FFmpeg filter, you need a different platform.

Can Very Good FFmpeg run custom FFmpeg filters?

Yes. Every FFmpeg filter and encoder is available. If it works in ffmpeg on the command line, it works in the API.

References

• Shotstack pricing: https://shotstack.io/pricing/
• Shotstack docs guide: https://shotstack.io/docs/guide/
• Shotstack API reference: https://shotstack.io/docs/api/
• Shotstack FFmpeg alternative page: https://shotstack.io/solutions/ffmpeg-alternative/
• Shotstack product page: https://shotstack.io/product/
• Shotstack vs Creatomate: https://shotstack.io/vs/creatomate-alternatives/
• Shotstack alternatives: https://shotstack.io/alternatives/
• Shotstack rendering benchmark: https://shotstack.io/learn/rendering-speeds-benchmark/
• Shotstack about us: https://shotstack.io/about-us/
• Shotstack GitHub: https://github.com/shotstack
• Shotstack HN search: https://hn.algolia.com/api/v1/search?query=shotstack&tags=story
• Very Good FFmpeg: https://verygoodffmpeg.com
• Very Good FFmpeg pricing: https://verygoodffmpeg.com/#pricing
• Very Good FFmpeg TypeScript SDK: https://www.npmjs.com/package/@verygoodffmpeg/sdk
• Very Good FFmpeg Python SDK: https://pypi.org/project/very-good-ffmpeg/