DeepL Reviews 689

I contacted DeepL regarding DeepL Voice for Meetings. I received an initial reply from their sales team, but after that the communication stopped. I followed up and sent a reminder, but never received any further response.

This is a concern from a customer perspective. If communication breaks down at the sales stage, it raises questions about reliability going forward.

I was not able to test DeepL Voice for Meetings at all, despite initial promises of a demonstration.

Separately, the core DeepL translator is useful for quick translation of short phrases and works well for that purpose.

Overall, a mixed experience: a solid core product, but disappointing and unreliable sales communication.

It was a good service, but recently it is a complete garbage. Slow, glitchy, most of the time doesn't even work without an account..

Worst customer support from Germany!! Never use again.

Was a great tool before. This AI stuff is killing it, and probably on purpose so users move back to the bigger names. Too bad.

Takes longer to translate than to open the dictionnary sometimes!

Do not use this crap. They keep abusing and disrespecting users. I used this app for many years but they have become extremely abusive. Just use AI and forget about it.

They claim that -500,000 characters are free-, but when I created an account and tried to use the API, nothing worked.

I then went back to their website and saw the pricing listed as -$5.49-, so I thought I would test it. I paid the fee and started using the **new API that was automatically created in my panel**, and it worked well.

However, in the middle of the night I received a $50 invoice.

It feels like something is wrong, and the company appears to be charging in a misleading way.

I'm from Ukraine and DeepL doesn't work in my region
in response support said that my territory is considered 'occupied'

HORRIBLE SERVICE, they added subscription with character count, I am paying for a month and in 3 days it is gone!

Deepl has made the workspace horrible for its users. So much space is now wasted on the screen for various menus that one does not need and there is a lot of blank space that serves no purpose whatsoever. Users only have a very tiny frame to actually work with the text (about three sentences at a time), as more than half of the screen is taken over by menus or blank space. Horrendously bad design, as if Deepl no longer cares whether one can actually get something done in Deepl. In the free version even this space is now half taken by a running banner with company logos! It seems that they are very actively trying to push users away and make their user experience with Deepl as miserable as possible, even if they pay a subscription. I have to work with long texts and for that it is imperative that I see as much text as possible, not only three sentences at a time!

Good translator, which has however introduced a lot of AI to "help" but which is just in the way.

Addittionaly, since 2026 they've also introduced a subscription model, with the need to buy a monthly subscription to translate beyond the limit of 500.000 characters per month. This is reached incredibly quickly, with for example this review alone is already well over 500 characters, showing how quickly this limit is reached. One thousand of these a month is not a lot.

Sorry, was perfect, now awkward.

(EN) I'm using "DeepL" for business. As a foreigner working in Germany it helps me to better communicate with the locals. The fact, that the platform was developed by germans gives it a benefit when translating to or from German. The function of adapting the translation (adjusting single words) is especially convenient for me. The business subscription has minimum advertisement and marketing communications are not annoying.

I signed up for a trial account and had set a reminder to cancel before being charged. Unfortunately, there was an overlap between two accounts. I successfully cancelled one, but missed cancelling the other, which resulted in being charged for an annual subscription that I do not need.

I contacted customer support immediately on the same day the annual charge was processed. I had only used the service once during the trial period and had not used it at all after the paid subscription began.

Despite this, my refund request was denied. The response stated that the subscription details and pricing were clearly presented during checkout and that completing the purchase required agreement to the Terms and Conditions.

While I understand company policies, refusing a refund on the very first day of a subscription, especially when the service has not been used, shows a lack of flexibility and customer goodwill.

Even if I require translation services in the future, this experience makes it unlikely that I will return. Policies matter, but how companies handle honest mistakes matters more.

Awful customer service. Buggy, broken platform. I am not even about to contact support because their contact page is broken. Terrible company, AVOID

Horrible AI translation and no refund. I tried using Deepl Ai translate for a Chinese document and it ended up being horrible, the translated text was off topic and unnatural. I requested a refund and was denied.

I have used deepL for a couple of weeks, I subscribed to a plan that costs about 10 euros a month. After just a few translations I realised the quality is really bad. A friend of mine who received a document asked me "this translation is really bad, do you need help?" She didnt know it was a robot-translation. I could accept 10 euros a month for good quality, but I have already reached the character limit for one month, and unlimited costs like 30 euros/month. Its not worth it, at all. I see others here also comment the quality has declined, obviosly because of more integrated "artificial intelligence" instead of human intelligence.

Premised that, in the few hours after activating the monthly subscription, I don't think I've translated or attempted to translate the 300,000 characters that are declared as the maximum limit. As they are saying me by asking ''for an update'' in ordere to work with the tool.

The pricing is excessive, and the restrictive monthly limits make it worse.

Any professional knows that workload fluctuates — some months are intense, others are lighter. Designing a service around the assumption that usage will be identical every month shows a complete misunderstanding of how professional work actually functions.

Imposing limits at this price point is already questionable. Making those limits rigidly monthly, with no flexibility or rollover, is simply indefensible. It is ridiculous.

For what is being charged, the structure feels fundamentally misaligned with the needs of the very professionals it claims to serve.

I cannot recommend this service.

I used it for a long time, but in the past half year or so it got progressively worse, by multiple overlays that must be clicked on, until someone can translate something.

And now one has to pay for the service, so it is back to other translators.

My wife was charged for an annual DeepL Pro subscription in December 2024 and 2025 that she did not knowingly authorize. DeepL claims a free trial was initiated and automatically renewed, but they were unable to provide verifiable proof of my consent (such as IP address or confirmed acceptance).

Despite refunding the identical charge from December 2025, DeepL refused to refund the 2024 charge, citing internal policy and elapsed time.

I recently escalated the matter through EU consumer and the protection channels, including the German Verbraucherzentrale, due to the lack of resolution.

I recently subscribed to DeepL Pro for use with Trados Studio 2024, for Chinese-to-English assignments. My decision was based on the fact that my client already relied on DeepL as well as the overwhelmingly positive online reviews. However, I soon discovered that most of these reviews actually referred to the free online version of DeepL, not the paid Pro subscription.

Unfortunately, the translations generated by DeepL Pro via the Trados API have been extremely disappointing—reminiscent of the early days of Google Translate, when output often made little sense and could even be unintentionally comical. In fact, the quality is not far removed from what the original web-based beta of Google Translate produced back in April 2006. I experimented with the only available setting—"Formal," with the options "Default," "Less," and "More"—but none of these made any noticeable difference in translation quality.

Here are a few minor but telling examples:
• "Venezuela" was translated as "Commission."
• "China-Venezuela" became "China and China," "China-Commission," or "Sino-Chinese."

While these errors can be manually corrected in short projects, they become unmanageable in larger projects spanning over 100 pages, making it more efficient to retranslate everything from scratch rather than rely on DeepL.

It’s extremely disappointing that a paid service delivers such poor results, especially when the free version is far superior. The whole point of using AI translation tools is to streamline the process, not to create more work.

On Monday, I plan to contact my DeepL account advisor in the hope that there’s a solution to improve the quality of the generated translations, or I will have to cancel my subscription.

Best,
Joaquim