Back in 2021 I was onboarding a German e-commerce client at Seahawk. Seven languages, WooCommerce, Elementor, and a deadline that was already slipping. I installed WPML, as you do, set up the language switcher, ran the first translation sync, and watched the Time to First Byte crawl past 2.8 seconds on a £40/month managed host. The client wasn't thrilled. Honestly, neither was I.
That was the moment I started actually testing other options instead of just assuming WPML was the answer. Four years and a few hundred multilingual projects later, I have opinions. Strong ones.
---
Why People Are Looking Past WPML in 2026
WPML isn't bad. Let me say that clearly. For complex editorial sites with a content team managing translations in a workflow, it's still genuinely capable. But there are three things that keep pushing clients (and me) toward alternatives.
Price creep. The Multilingual CMS plan, the one you actually need for WooCommerce or custom post types, sits at $99/year now. That's per site. Agencies doing white-label work for SMEs can't always pass that on without a conversation they'd rather not have.
Performance overhead. WPML adds a non-trivial number of database queries per page load. On a large site with thousands of posts, that compounds. I've seen it add 400-600ms to TTFB on sites that weren't tuned specifically around it. Google's Core Web Vitals guidance makes that kind of drag increasingly expensive to ignore.
Complexity for simple needs. If a client needs three languages and twenty pages, WPML's configuration surface is intimidating. You're clicking through String Translation, Translation Management, WooCommerce Multilingual setup... it's a lot. Sometimes you just need something that works in an afternoon.
---
Polylang: The Workhorse Most Agencies Sleep On
I've used Polylang on probably 200+ sites at this point. The free version handles basic post/page/taxonomy translation without charging you a penny. The Pro version (£99/year for unlimited sites, last I checked) is where it gets genuinely competitive, because that "unlimited sites" licensing is a completely different proposition for agencies compared to WPML's per-site model.
What Polylang Does Well
The UI is clean. Language assignment happens right in the post editor sidebar, it doesn't take over your dashboard, and the performance impact is noticeably lighter than WPML, on a like-for-like test I ran in March 2025 on a staging site with 500 posts and WPML vs Polylang Pro, Polylang shaved around 180ms off TTFB. Not enormous, but real.
It integrates fine with Elementor, Bricks, ACF, and WooCommerce (via the WooCommerce add-on, which costs extra). That last bit is the catch: WooCommerce support needs the add-on, and once you stack that in, the cost difference with WPML narrows a bit.
Where Polylang falls down is machine translation. There's no built-in DeepL or Google Translate integration in core. You need a third-party bridge like WPML's own ecosystem or something like Linguise, which brings me to...
---
TranslatePress: The Best Option for Speed-First Projects
TranslatePress changed how I think about multilingual WordPress. Instead of the post-duplication model (where you have separate posts per language), it uses a front-end visual editor where you translate text directly on the rendered page. You see your site, you click a string, you type the translation. Done.
This matters for performance. There's no language-specific post duplication inflating your database. Translated strings live in a custom table, not as separate posts. On a project Seahawk ran for a French-Belgian retail client in late 2024, we switched from WPML to TranslatePress mid-project (not ideal, I know, but the client changed scope) and the database query count per page dropped by about 40%.
The DeepL Integration
TranslatePress Pro has a DeepL API integration built in. For clients who want machine translation as a starting point and human review after, this workflow is miles better than anything I've patched together manually. You set it to auto-translate on publish, your editor reviews and corrects, and you're done. The turnaround on a 50-page site goes from days to an afternoon.
Pricing sits at around $89/year for the Personal plan, $179 for Business (3 sites), $299 for Developer (unlimited). For agencies the Developer plan is the one to look at.
One honest limitation: very complex WooCommerce stores with lots of dynamic content can expose gaps in string detection. TranslatePress catches most things, but if you have heavily customised checkout flows or plugin-generated strings from obscure add-ons, you might miss some. Budget a few hours for manual string mopping-up.
---
Weglot: When the Client Has Budget and Wants Zero Friction
I'll be direct: Weglot is expensive. £99/year gets you one language pair and up to 10,000 words. A real business site blows past that almost immediately, and the pricing scales up quickly from there. For a mid-size site with three languages you could be looking at £300-400/year.
But here's the thing. Setup is about 15 minutes. You install the plugin, paste an API key, pick your languages, and it works. No configuration rabbit holes, no compatibility headaches with your page builder. It handles JavaScript-rendered content better than any other option on this list, which matters if you're running a React-heavy or Elementor-with-lots-of-dynamic-widgets setup.
I recommend Weglot specifically when: the client is a startup with real revenue, they need to move fast, and multilingual is genuinely peripheral to their product (not the main event). For a SaaS landing page going from English to German and French, it's a no-brainer if budget isn't the objection.
---
MultilingualPress: The One for WordPress Multisite Setups
If you're running WordPress Multisite, and in 2026, multisite is having a bit of a quiet resurgence for enterprise and franchise clients, MultilingualPress is the plugin you want. Full stop.
It creates a separate site per language within your network. This is a fundamentally different architectural approach compared to the others on this list. Each language site is independent: its own database records, its own URL structure (you can do subdomains or separate domains), its own admin. They're connected by relationships that let you link equivalent content across sites.
Seahawk had a franchise client in 2023 with 14 regional sites, 6 languages, and a head office that wanted central content control with local editorial override. MultilingualPress on a multisite network was genuinely the right answer. Nothing else came close. MultilingualPress's own documentation is actually decent if you want to see how the content relationships work.
The downside is operational complexity. Multisite is harder to maintain, harder to hand off to a non-technical client, and plugins that don't support multisite (there are still a few) become your problem. Go in with eyes open.
---
Linguise: The Dark Horse for Headless and API-Driven Projects
Linguise isn't a WordPress plugin in the traditional sense. It's a proxy-based translation layer that sits in front of your site and handles translation at the edge. This makes it unusually interesting for headless WordPress setups where the frontend is built in Next.js or Nuxt.
I tested it on a headless project in early 2025 where the client had a Gutenberg-based CMS feeding a Next.js frontend. WPML and TranslatePress were both awkward fits because translation needed to happen in the rendered output, not in WordPress content fields. Linguise handled it cleanly.
It's not cheap either, plans start around €15/month, but for the right architecture, nothing else really solves the problem the same way.
---
How to Actually Choose
Here's my decision framework after building multilingual sites across twelve thousand projects. Not a rigid algorithm, but a set of questions that narrow it down fast.
- What's the URL strategy? If the client needs separate domains per language (
example.de,example.fr), MultilingualPress or WPML handle this most cleanly. TranslatePress and Polylang can do it but need more config. - Is WooCommerce involved? If yes, shortlist WPML, Polylang Pro + WooCommerce add-on, or TranslatePress Business. Test each on your specific product catalogue size.
- Does the client want machine translation? TranslatePress with DeepL or Weglot are the cleanest. WPML's translation management also works but feels more enterprise-y.
- How many sites does your agency need to cover? Polylang Pro's unlimited-sites licence is the standout value for agencies at volume.
- Is the frontend headless or heavily JavaScript-driven? Weglot or Linguise. Don't fight with post-duplication plugins on a Next.js frontend.
- Is this a multisite network? MultilingualPress. That's it.
---
What About Just Using the Block Editor's Native Multilingual Features?
WordPress core has been inching toward native multilingual support for years. Matt Mullenweg has referenced it publicly more than once, and there have been Gutenberg GitHub discussions about it going back to 2020. As of early 2026, it still isn't here in any meaningful production-ready form.
I wouldn't hold off on a plugin decision waiting for it. Build with what exists today. If native support lands, migrating is a problem you'd welcome.
---
FAQ
Is WPML still worth using in 2026?
For large editorial sites with dedicated translation teams, absolutely. The Translation Management module, the XLIFF export, the compatibility with almost everything in the WordPress ecosystem, it's still the most complete option. Where it loses is price at scale, performance on large databases, and complexity for simple projects.
Can I switch from WPML to another plugin mid-project?
Technically yes, practically it's painful. WPML stores language metadata in its own tables and post meta. Most other plugins don't read that natively. There are migration tools and scripts floating around on GitHub, but expect to spend time on it. I've done three migrations off WPML and every one needed custom SQL work. Plan for half a day minimum on a medium-size site.
Does TranslatePress work with Elementor and Bricks?
Elementor: yes, very well. TranslatePress's front-end translator renders the Elementor output and captures strings reliably. Bricks is more recent and the compatibility has improved through 2024-2025, I'd run a thorough test on your specific template before committing on a production site.
Which multilingual plugin is best for SEO?
All the major ones (WPML, Polylang, TranslatePress) implement hreflang tags correctly when configured properly. The gap isn't in hreflang implementation, it's in how fast each renders the page. SEO in 2026 is heavily weighted toward Core Web Vitals, so the performance difference between a bloated WPML install and a lean TranslatePress setup can matter more than the hreflang configuration.
What's the cheapest way to do multilingual WordPress?
Polylang free plus manual translation for each post. Zero cost. Works fine for small sites with patient editors. The moment you need WooCommerce, machine translation, or more than 50 posts, you're going to want to pay for something.
---
The multilingual WordPress space has genuinely matured. WPML had the category to itself for too long, and competition made everything better. Pick based on your architecture, your client's budget, and your own willingness to configure things, not based on what was the default answer in 2018.