Subtitle Translation for Netflix and Amazon: How to Start in 2026

$5 per minute of video - that’s the average rate for translating subtitles for streaming platforms. One 45-minute episode = $225. Two episodes a day = $450. Sounds like a freelancer’s dream, right? Here’s the catch: one minute of video takes 15-20 minutes of actual work, getting into Netflix’s translator pool involves a five-stage test with tough pass rates, and AI-generated subtitles are eating into the market fast. Let’s break down how this niche actually works, what you can realistically earn, and how to get started.

Why video localization is a growing market¶

Streaming platforms are creating massive demand for subtitles. Netflix operates in 190+ countries and translates content into 30+ languages. Amazon Prime Video, Disney+, HBO Max - they’re all expanding globally. The global dubbing and subtitling market hit $13.1 billion in 2024 and keeps growing.

Subtitling isn’t just “typing words under a video.” You need to work with timecodes, stay within character-per-second limits, account for viewer reading speed, and sync text with audio and scene changes. It’s a separate specialization with its own rules and tools - and that’s exactly why it pays better than general translation.

How the industry works: from Netflix to YouTube¶

Netflix: the Hermes test and vendor model¶

Netflix doesn’t hire subtitlers directly. The platform works through Preferred Vendors (NPVs) - major localization companies: Iyuno (formerly Iyuno-SDI Group), ZOO Digital, Keywords Studios. These companies get orders from Netflix and distribute work to freelancers.

To get into the pool, you need to pass Hermes - Netflix’s five-part test. The first four sections are multiple choice (grammar, language, cultural context), and the last two are practical subtitling tasks. The test takes 90 minutes, and the pass score depends on the language pair: Italian and Portuguese require about 90%, while less common languages have lower thresholds.

Here’s the thing: as of 2026, Netflix has filled its translator pool for most language pairs and periodically closes Hermes registration. But the path isn’t blocked - you can apply directly to vendors (Iyuno, ZOO Digital, Keywords Studios), which have their own tests and constantly recruit freelancers.

Amazon Prime Video: PLV program¶

Amazon works through Preferred Localization Vendors (PLVs) - large language service providers. The model mirrors Netflix: freelancer works with vendor, vendor works with Amazon. Amazon has its own Language-Specific Style Guides with technical subtitle requirements.

YouTube: the easiest entry point¶

YouTube is where you should start. Thousands of creators are looking for subtitle translators. Rates are lower, but requirements are more relaxed. You can reach out to YouTubers directly, work through platforms like Rev or HappyScribe, or find gigs on Upwork and Fiverr.

YouTube subtitling is the perfect training ground. You can practice timing, build speed, and put together a portfolio for bigger platforms.

Technical requirements: what you need to know¶

Subtitle formats¶

• SRT (SubRip) - the most common format. A simple text file with timecodes. 90% of jobs come in SRT.
• ASS/SSA (SubStation Alpha) - supports styling, colors, fonts. Used in anime and fansubbing.
• VTT (WebVTT) - web standard, used on YouTube and in browsers.
• TTML/DFXP - the format used by Netflix and other streaming platforms.

Timing rules (spotting)¶

This is the most critical technical skill for a subtitler:

• Reading speed: 12-17 characters per second (CPS) is the industry standard. Netflix allows up to 20 CPS for adult content and 17 for children’s.
• Max characters per line: 35-42, usually no more than two lines at once.
• Minimum subtitle duration: 1 second (or 15 frames).
• Maximum duration: 7-8 seconds.
• Sync: subtitle appears when speech starts and disappears when the phrase ends.
• Shot changes (cut changes): subtitles shouldn’t “hang” across a hard cut between scenes.

Text expansion in translation¶

English text expands by about 30% when translated to German, 25% to Spanish, and 20% to French. This means subtitlers often need to rephrase, condense, and adapt text to fit the same timecodes while preserving meaning - all within strict CPS and character-per-line constraints.

Subtitler’s toolkit¶

Free tools¶

• Aegisub - the most popular free subtitle editor. Powerful styling, audio waveform-based timing, ASS/SSA support. Great for learning and YouTube work.
• Subtitle Edit - supports 200+ formats, format conversion, built-in timing and CPS verification. Very handy for SRT work.

Professional tools¶

• Ooona - the cloud platform used by Netflix vendors. This is where most professional streaming subtitling happens.
• EZTitles - professional subtitling and format conversion software.
• CaptionHub - cloud-based platform for team subtitling projects.

For getting started, Aegisub or Subtitle Edit is all you need - both are free and cover everything. When you start working with vendors, they’ll give you access to their own tools (usually Ooona or similar).

How much do subtitlers earn¶

Let’s look at real numbers.

Rates per minute of video¶

Monthly income breakdown¶

Realistic calculation for a mid-level subtitler:

• 1 minute of video = roughly 15-20 minutes of work (translation + timing + QC)
• In an 8-hour workday, you can realistically process 25-30 minutes of video
• At $5/minute: 30 x $5 = $150/day
• 20 working days = $3,000/month

A beginner on YouTube at $2-3/minute with slower output will earn $800-1,200/month. An experienced subtitler with steady work from Netflix vendors - $4,000-5,000.

On the ProZ.com forum, one translator noted that $3/minute is “not fair, but common.” They recommend not working below $5/minute and aiming for $8. Just like with medical translation, specialization directly impacts your rates.

How to break in: step-by-step plan¶

1. Learn the fundamentals¶

Subtitling isn’t just translation with timestamps. You need to understand spotting rules, CPS limits, file formats, and sync principles. Start with Netflix’s Timed Text Style Guide - it’s publicly available at partnerhelp.netflixstudios.com and is the industry gold standard.

2. Master the tools¶

Download Aegisub or Subtitle Edit (both free). Grab any YouTube video and practice:

• Creating subtitles from scratch (transcription + timing)
• Translating existing subtitles while preserving timecodes
• Checking CPS and spotting rule compliance

Don’t rush - your first few videos will be slow. That’s normal, speed comes with practice.

3. Build a portfolio¶

• Translate subtitles for 3-5 videos across different genres (documentary, series, interview)
• Show you can work with multiple formats (SRT, VTT)
• Offer free translation for a small YouTube channel - that’s real experience plus a reference

Some of the most successful subtitlers started with fansubbing - translating anime subtitles for fan communities. It’s unpaid work, but it gives you real practice and connections.

4. Start with accessible platforms¶

• Rev - requires a test and a work sample. Lower rates, but plenty of orders.
• HappyScribe - assessment test on signup. Solid platform for beginners.
• Upwork / Fiverr - search for “subtitle translation” gigs. Competition is high, but you’ll build experience fast.

5. Apply to vendors¶

Once you have 6-12 months of experience and a portfolio:

• Iyuno - register through iMTalent (their freelancer platform). Tests, training, and access to Netflix, Amazon, Disney+ projects.
• Keywords Studios - join their Talent Community as a freelance subtitler.
• ZOO Digital - a Netflix NPV that recruits freelancers for various language pairs.

6. Take the Hermes test (if registration is open)¶

Netflix’s Hermes test opens periodically for new language pairs or when the pool needs refreshing. Keep an eye on tests.hermes.nflx.io. If the test is closed, work through vendors - the end result is the same.

AI and subtitling: what’s changing¶

Let’s be honest: AI is hitting the subtitling industry hard. A 2024 Acolad survey found that 84% of translators predict declining demand for human translation due to AI. Whisper, ElevenLabs, and specialized AI subtitling tools (Moovly, Subtly, DeskDub) already generate first-draft subtitles automatically.

But auto-generated subtitles are a draft, not a finished product. AI struggles with:

• Slang, dialects, and accents
• Cultural references and humor (try explaining to a machine why “I see dead people” needs different treatment in a comedy vs. a horror film)
• Precise timing around shot changes
• SDH (Subtitles for the Deaf and Hard of Hearing), where sounds and music need to be described

What this means for you: translating subtitles from scratch is a dying format. MTPE for subtitles is the new reality. A subtitler who can quickly clean up an AI draft, fix timing issues, and add cultural adaptation - that’s exactly who vendors are hiring. Output goes up, quality stays the same.

Don’t fear AI - learn to work with it. Use ChatGPT or Claude for a rough first draft, then polish in Aegisub. This hybrid approach is already becoming the industry standard.

FAQ¶

How long does it take to subtitle one TV episode?¶

A standard 45-minute episode takes 8-12 hours from scratch (translation + timing + QC). With MTPE (editing an AI draft) - 4-6 hours. The more experience you have and the better you know the genre, the faster you’ll work.

Which language pairs are most in demand?¶

Korean, Japanese, Arabic, Danish, and Swedish are languages where Netflix is actively looking for translators. Overall, any non-dominant language pair where demand outstrips supply gives you an advantage. If you work with less common languages, your chances of getting steady vendor work are significantly higher.

Do I need a degree in translation?¶

No. A linguistics or translation degree is a plus, but not a requirement. What matters more: excellent command of your language pair, understanding spotting rules, and proficiency with subtitling tools. Some of the most successful subtitlers came from fansubbing with no formal education.

Can I combine subtitling with regular document translation?¶

Yes, and many translators do. Subtitling requires different skills (timing, text condensation), so it complements document translation nicely. If you work on documents too - ChatsControl can handle the document side quickly while you focus on audiovisual projects. Just keep in mind that Netflix and Amazon vendor work often comes with tight deadlines, so you’ll need schedule flexibility.

Will AI replace subtitlers?¶

It already is for simple content (corporate videos, training courses). But for cinematic content, TV series, and comedy - human adaptation is still critical. Translators who combine subtitling skills with AI proficiency have the best outlook. MTPE for subtitles lets you process twice the material in the same time.