Netflix Subtitle Guidelines Explained

The document that quietly became the industry’s shared definition of a well-made subtitle — and the exact numbers behind it.

Netflix’s Timed Text Style Guide (TTSG) is one of the most detailed, publicly available subtitle specifications in the streaming industry, and it has become something close to a de facto standard well beyond Netflix’s own platform. Reading-speed limits, line-length rules, and timing conventions that originated in Netflix’s internal specification now show up, in close to identical form, across other streamers, YouTube style guides, and independent creators’ subtitle workflows — not because those platforms coordinated with Netflix, but because the guide represents some of the most rigorously tested public thinking on what makes a subtitle actually readable.

This breakdown walks through what the TTSG actually requires — file format, reading speed, character limits, timing, positioning, and SDH — and explains why so much of it has been adopted well outside Netflix’s own delivery pipeline. Whether the goal is delivering to Netflix directly, or simply borrowing from the most tested subtitle standard available, understanding these rules in detail is useful for almost anyone producing subtitles professionally.

Why Netflix’s Guide Became the Industry Benchmark

Netflix publishes its Timed Text Style Guide openly through its Partner Help Center, broken into a general requirements document plus dozens of language-specific guides covering grammar, punctuation, and formality conventions unique to each language. That level of public detail and language-by-language granularity is unusual — most streaming platforms manage subtitle specifications through closed vendor relationships rather than a searchable public knowledge base — and it’s a major reason the TTSG gets referenced, adapted, and quoted well outside Netflix’s own vendor network.

The guide is also unusually rigorous about the mechanics of readability specifically. Its reading-speed and timing figures didn’t emerge from guesswork; they reflect extensive testing on how quickly viewers can actually process on-screen text without falling behind or losing track of the picture. That’s the underlying reason so many of these numbers keep reappearing verbatim across other platforms’ specifications and independent style guides.

File Format and Technical Requirements

Netflix’s subtitle delivery format is built on TTML (Timed Text Markup Language), with strict rules about how styling is expressed inside the file. Two technical details stand out because they trip up files built for other platforms:

• No pixel values, anywhere. Positioning uses tts:textAlign and tts:displayAlign with static values for tts:extent and tts:origin, and font size must be defined as a percentage — specifically 100% — rather than a fixed pixel size. This keeps subtitles resolution-independent across every device and screen size Netflix supports.
• A restricted glyph list. Only characters included in Netflix’s approved Glyph List can be used in a delivered file — an easy detail to miss when a source file was built in a general-purpose subtitle editor that doesn’t check against Netflix’s specific character set.
• Arial is used as the placeholder reference for a generic proportional sans-serif font, with actual on-screen font size relative to video resolution and calculated to reliably fit 42 characters across the screen — not a fixed number tied to any particular resolution.

For teams working across multiple delivery formats, not just Netflix’s, our SRT vs VTT vs ASS subtitle format guide is a useful companion to this section, since it covers how the more common general-purpose subtitle formats differ from the strict TTML-based delivery format Netflix specifically requires.

Reading Speed and Timing Rules

This is the part of the TTSG most frequently cited outside Netflix, because the numbers have become something close to an unofficial industry default:

Files that exceed these thresholds get flagged automatically during Netflix’s quality-control process, before a human reviewer ever looks at the content — meaning a technically accurate translation with a reading-speed violation still fails delivery. The standard fix for a CPS violation is to split the line or trim the wording, not to speed up the timing, since compressing the display duration further only makes the reading-speed problem worse.

Character Limits and Segmentation

Netflix’s 42-character-per-line limit is probably the single most widely copied number from the entire guide — it shows up, often unchanged, in YouTube’s own subtitle recommendations and in countless independent creator style guides. Text should generally be kept to one line unless it exceeds that limit, and when a line does need to break into two, segmentation has to happen at a clause level so each line reads as a complete, logically self-contained thought rather than an arbitrary word-count split.

For subtitles split across two consecutive events because a sentence continues, timing and punctuation both follow specific rules: ellipses are generally not used when the pause between the two parts is under two seconds, but are used when the pause is longer or when dialogue trails off. Netflix additionally specifies the single smart ellipsis character (U+2026) rather than three separate periods — a small detail, but one that automated QC checks for directly.

Punctuation Conventions Differ Even Within English

One detail that surprises people encountering the TTSG for the first time: Netflix maintains separate style guides for different English variants, and they don’t always agree with each other. US English uses double hyphens to indicate an abrupt interruption by another speaker or a sudden sound; UK English uses an ellipsis for the same situation instead of hyphens. Getting this backward — using US conventions on a UK-English delivery, or vice versa — is exactly the kind of detail that looks like a stylistic quirk but is actually a hard specification difference that automated and human QC both check for.

This regional specificity extends throughout the guide: each language-specific TTSG covers its own conventions for quotation marks, italics, numbers, currency, and formality — currency mentioned in dialogue, for instance, is kept in its original currency rather than converted, regardless of the target language.

Positioning Rules

Subtitles are required to be center-justified and placed at either the top or bottom of the screen — Netflix doesn’t support left- or right-aligned subtitle blocks for standard delivery. The one documented exception is Japanese content, where vertical positioning is permitted under Netflix’s Japanese-specific Timed Text Style Guide, reflecting the vertical reading conventions native to the language.

When on-screen text (like a sign or a letter shown in the frame) overlaps with spoken dialogue, Netflix’s guidance is to prioritize whichever message is more plot-pertinent, and specifically warns against over-truncating dialogue or severely reducing reading speed purely to cram in both the on-screen text translation and the spoken dialogue at once.

SDH: Subtitles for the Deaf and Hard of Hearing

Netflix treats SDH as a distinct deliverable from standard subtitles, not a lightly modified version of the same file. SDH includes non-dialogue audio information — sound effects, music cues, tone, and speaker identification — that a standard subtitle track deliberately omits, on the assumption that a standard-subtitle viewer can already hear that information.

For a broader look at what separates SDH from ordinary subtitles and why the distinction matters beyond Netflix specifically, our guide to making video content deaf-friendly and our ultimate guide to video accessibility both cover the accessibility side of this in more depth.

Translation Consistency: KNP and Formality Tables

For episodic content, Netflix requires delivery teams to create and maintain Key Name and Phrase (KNP) tables and formality tables — reference documents that lock down how character names, recurring terms, and levels of formal or informal address are translated, and keep that translation consistent across every episode and season. Without this, it’s easy for a long-running series translated by different subtitlers over time to drift — a character’s name spelled two different ways, or a formal-address convention that flips partway through a season.

Translator credit requirements are similarly specific: subtitle translators are generally credited on-screen (unless they’ve formally waived that right), the credit can’t appear on screen at the same time as the Netflix ident, and if more than one translator worked on an asset — say, across multiple source languages or a collaborative project — all contributing translators can be named in the same credit.

Pivot Templates for Unusual Language Pairs

When a title needs to move between two languages with no direct, readily available translator pool — Japanese to Polish, for example — Netflix’s process allows for a pivot template: a fully timed and edited subtitle file in an intermediate language (often English) that bridges the source and target languages. Because this adds an extra layer of language transfer, the guide is explicit that the pivot file has to function with the same rigor as a direct source-to-target translation would, rather than being treated as a rougher intermediate step.

Why These Numbers Show Up Everywhere Else

It’s worth being direct about why a private company’s internal delivery specification ended up shaping subtitle practice this broadly: the TTSG is public, detailed, and grounded in real viewer-comprehension testing, at a time when very few other platforms published anything comparable. A 17 CPS ceiling, a 42-character line limit, and a 7-second maximum duration are specific, testable numbers — easy to adopt wholesale rather than independently re-deriving similar limits from scratch. The result is that Netflix’s specification functions as a shared reference point across the industry even for creators and platforms with no delivery relationship to Netflix at all.

Working Toward Netflix-Grade Subtitles, Even Outside Netflix

Most creators and businesses will never deliver a file directly to Netflix’s Partner Help Center, but building subtitles against the same discipline the TTSG encodes — accurate reading speed, clean segmentation, consistent terminology, proper SDH — produces a noticeably higher-quality result on any platform:

1. Start with an accurate, reviewed transcript rather than a raw automatic one, since every downstream timing and character-limit decision depends on the underlying text being correct.
2. Time subtitles to the same reading-speed discipline Netflix enforces — roughly 17 CPS for adult content, 13 CPS for children’s content — even on platforms that don’t check for it automatically.
3. Keep lines to 42 characters where possible, breaking at clause boundaries rather than arbitrary word counts, and defaulting to one line unless a second is genuinely needed.
4. Build a simple KNP-style reference sheet for any multi-episode or recurring content, locking down character names and key terms before translation begins rather than fixing inconsistencies after the fact.
5. Produce a full SDH version alongside standard subtitles wherever accessibility matters — not a stripped-down variant, but a genuinely separate deliverable with sound and speaker information included.

vSubtitle supports this workflow directly: its AI subtitle generator produces accurate, editable captions with reading-speed guidance built in, and exports to the SRT, VTT, TXT, and DFXP formats most commonly used as the starting point for further platform-specific conforming. For teams just getting oriented, our beginner’s guide to AI subtitling and our roundup of the best free subtitle generators are good starting points before tackling a stricter spec like Netflix’s.

Key Takeaways

• Netflix’s Timed Text Style Guide is unusually detailed and publicly available, which is a major reason its reading-speed and formatting rules have spread well beyond Netflix’s own delivery pipeline.
• Reading speed is capped at roughly 17 CPS for adult content and 13 CPS for children’s content, with a 5/6-second minimum and 7-second maximum duration per subtitle event.
• The widely copied 42-character-per-line limit, along with clause-level segmentation, is central to how Netflix defines a readable subtitle.
• Technical delivery requires percentage-based, not pixel-based, sizing and positioning, plus adherence to Netflix’s specific approved glyph list.
• Even punctuation conventions vary by language and region — US and UK English use different marks for the same kind of interruption.
• SDH is a distinct deliverable from standard subtitles, and episodic content requires KNP and formality tables to keep translation consistent across a series.

Whether or not a delivery to Netflix itself is on the table, the TTSG remains one of the clearest, most rigorously tested references available for what a well-made subtitle actually looks like — which is exactly why so much of the rest of the industry keeps building toward the same numbers.

Frequently Asked Questions (FAQs)