Lifestyle
Cloudflare open-sources Streamline: a demo of using its cloud services to add graphics to live streams and burn subtitles into videos
On October 2, 2026, Cloudflare launched and open-sourced Streamline, a developer playground showing how developers can combine Stream, Workers, Containers and Durable Objects to build their own video processing pipelines, such as adding graphics to live streams in real time or adding subtitles to videos. This article explains what it is, how it works, its limitations, and what it means for viewers and developers.
About 7 min read
What Cloudflare announced
Cloudflare published a blog post on October 2, 2026, written by Willi Geiger and Taylor Smith. In it, the company says it has launched a developer playground called Streamline. Cloudflare notes that its Stream platform (Cloudflare's service for live streaming and video hosting) "just works" for many customers. Some tasks need more, such as drawing dynamic annotations on a live stream in real time, or producing a version of an uploaded video with burned-in subtitles (subtitles composited directly into the picture). These require a custom video processing pipeline, meaning a sequence of steps that process video in order. Streamline was designed to demonstrate exactly these kinds of needs.
Cloudflare says it has released Streamline as open source (publicly available code that anyone can view and use) on its GitHub. The code is split into two repositories: the media engine, cloudflare/streamline, and the demo app, cloudflare/streamline-demo. Cloudflare has also deployed a public demo for developers to try.
How Streamline works
According to Cloudflare, the system has two main parts. A "media engine" handles video input, output and processing. A "control app" creates, configures, observes and stops processing sessions (a session is a single video processing job). The media engine runs in Containers, which Cloudflare describes as a long-running execution environment suited to media processing. Inside the engine are a controller written in Go and a processor that does the actual video work. Cloudflare notes the processor currently uses FFmpeg, but says this is an internal implementation detail. The control app is built on Workers (Cloudflare's cloud code execution service, used here to send control commands and to monitor). Durable Objects, a Cloudflare component that helps with orchestration, coordinates sessions, container lifecycles and preview relaying.
Cloudflare says the media engine can pull a live stream from one Stream Live input over RTMPS (a common live streaming transport protocol) and output it to another Stream Live input. It can also read Cloudflare Stream HLS (HTTP live streaming) manifests to use a hosted video as a source, or receive webcam video sent by the control app. The processed result can be output as a live stream. It can also be sent back as fMP4 video segments over WebSocket (a technology that keeps an app connected to a server for continuous data transfer) to give a low-latency preview.
- filter: apply filters, such as blur or saturation
- overlay: overlay images, including PNG annotation layers that can be updated while running
- subtitle: burn subtitles into the picture
- encode: set output encoding parameters
Cloudflare adds that the engine currently fixes the order in which these operations run. Another key design point is that once a session has started, the controlling app can disconnect and reconnect safely, and the container keeps processing until it is stopped. The system also has a maximum run time, so sessions cannot run indefinitely.
Comparing the three operating modes
| Mode | Access control | Session limits | Cloudflare's description |
|---|---|---|---|
| Local development | No authorization needed for local access | Single user | Container is local Docker, no Durable Object is used, and preview connects directly to a local WebSocket |
| Owner deployment | Kept private through Workers' Access integration | Only one session at a time; others cannot stop or replace it | Deliberately private; not a security model for a public multi-user service |
| Public demo | Uses its own Access configuration | One container identity and one active session per verified user | Has global admission control, plus concurrency, media and session limits |
Security design and current limitations
Cloudflare says security is part of the design. Only authorized users can create or take over sessions, sessions are isolated from one another, and resource usage is bounded. According to the company, Stream Live Input keys are stored either in Worker secrets or as write-only shared overrides in Durable Object storage. They are never returned by the settings API and are never placed in browser storage.
On limitations, Cloudflare admits the current version processes media on Container CPUs, which becomes a bottleneck at higher resolutions or frame rates. The company says it hopes to extend the architecture with the developer community. Its goals include computer vision pipelines, hardware-accelerated processing, and next-generation real-time protocols such as WebRTC and MoQ, and ultimately native video encoding and decoding in Workers.
What it means for general readers and developers
For ordinary viewers, tools like this could mean more live streams and online videos with real-time processing, such as live subtitles, dynamic graphics or picture-in-picture. Cloudflare also mentions other uses, such as sending factory camera feeds for AI analysis or combining several camera feeds into a single picture. However, this is still a demo project. Whether it catches on depends on developer adoption and future performance improvements.
For developers, according to Cloudflare, the demo app uses an Astro front end to showcase overlays, subtitle decoding, filters and picture-in-picture. It also provides probes for performance metrics and system tracing. The app can run locally, or it can be deployed behind Cloudflare Access to control who can use it. Anyone interested can read the original post first to understand the architecture and limitations, then decide whether it fits their needs.
Cloudflare official blog: original Streamline post
FAQ
Is Streamline an official commercial product?
According to Cloudflare, Streamline is a "developer playground" that demonstrates how to build custom video pipelines on its developer platform, and it is released as open source. The original post does not describe it as a standalone paid product.
What video processing can Streamline do?
According to Cloudflare, it currently supports four types of operations: filters, image overlays (including PNG annotations that can be updated while running), burned-in subtitles, and output encoding settings. The engine fixes the order in which they run.
Do ordinary viewers need to do anything?
No. This is a tool for developers, and ordinary viewers need not take any action. Its impact is more likely to appear indirectly in future live streaming or video services, such as live subtitles or dynamic graphics.
What are Streamline's known limitations?
Cloudflare admits the current version processes media on Container CPUs, which becomes a bottleneck at higher resolutions or frame rates. The company also says the owner deployment can run only one session at a time and is not a security model for a public multi-user service.
Where does this information come from?
Everything in this article comes from Cloudflare's official blog post published on October 2, 2026, and reflects the company's own account.
Browse the latest news in this topic
Lifestyle
Google unveils Gemini 4 Argon: cyber defenders get it first, everyone else still has to wait
On September 30, 2026, Google announced Gemini 4 Argon, which it calls its new frontier (most advanced) AI model. For now it is available only to trusted cyber defenders through the Fairwind Program. Here is what Google says the model can do, what it will cost developers, how Google says it is managing the risks, and what it means for everyday users. All figures come from Google itself.
Lifestyle
NVIDIA: CoreWeave Begins Offering Vera Rubin NVL72, With Cognition as First Production Customer
According to the NVIDIA blog, AI cloud provider CoreWeave now offers NVIDIA's next-generation Vera Rubin NVL72 systems, plans to offer the Vera CPU and launched CoreWeave Forge. This matters mainly to companies building AI agents, and could eventually mean faster AI tools for everyday users.
Lifestyle
Google Cloud makes Spanner Omni generally available: its Spanner database can now run in companies' own data centers and on other clouds
Google Cloud says Spanner Omni, a version of its Spanner database that businesses run themselves, is now ready for real-world use in their own data centers, on other clouds or on a laptop. This matters to organizations that want Spanner outside Google Cloud, but they take on the running of it. Here are the features, licences and trade-offs Google Cloud describes.
Lifestyle
Google Cloud Previews a Remote MCP Server That Lets AI Agents Run gcloud and bq Commands for You
Google Cloud has opened a public preview of its Google Cloud CLI remote MCP server. It lets AI agents run Google Cloud's gcloud and bq command-line tools in isolated cloud sandboxes, with nothing to install locally. This article explains how it works, the safeguards Google describes, the pricing, and what it means for businesses, developers and everyday readers.
Latest travel guides
GuideTokyo
Where to Stay in Tokyo: Comparing Shinjuku, Ueno, Tokyo Station, Shibuya, Asakusa, Ikebukuro, and Ginza, Plus Airport Access, Accommodation Tax, and Luggage Delivery
Where should you stay in Tokyo? Compare Shinjuku, Ueno, Tokyo Station, Shibuya, Asakusa, Ikebukuro, and Ginza by the same criteria: access from Narita and Haneda, transit routes, nearby attractions, neighborhood character, and who each area suits. Includes a comparison table, a Yamanote Line diagram, Tokyo’s accommodation tax as verified in 2026/9 (changing to 3% in 2027/4), and Airport TA-Q-BIN luggage shipping rules.
- Budget
- Hotels
GuideTokyo
How to Choose Tokyo Transit Passes: Are Suica, Welcome Suica, the Tokyo Subway Ticket, and the JR Pass Worth It?
On a first Tokyo trip, start with an IC card and pay per ride (Welcome Suica has no deposit and is valid for 28 days). If you take four or more subway rides in a day, add a 72-hour Tokyo Subway Ticket for 2,000 yen; a JR Pass is never worthwhile if you stay in Tokyo and do not go to Kansai. See what TOURIST PASMO, Suica on iPhone, and the Tokyo Metro day pass do and do not cover, with a decision chart. Prices verified in September 2026.
- Transport
- Budget
GuideTokyo
Tokyo Disneyland and DisneySea Guide: Ticket Prices, Fantasy Springs, Disney Premier Access (DPA), Standby Pass, and Which Park to Choose for Your First Visit
Tokyo Disney one-day Passport prices vary: most weekdays in 9/2026 cost ¥9,900 and weekends ¥10,900. At 14:00 daily, tickets go on sale for the same date two months later. Free Priority Pass is no longer on the official service list; only paid Disney Premier Access (¥1,000–3,500 per person per use) shortens waits. Covers hours, the 25th anniversary, Standby Pass, Entry Request, Fantasy Springs access and first-visit park choice; checked on the official site in 9/2026.
- Itineraries
- Family