Dialora Reviews

3.6

3.65 stars
3.65 stars
57

Taco ratings

29
3
8
10
7
AskSumo

The review summary should be corrected to: "Customers are praising Dialora for its efficient setup, responsive support, and reliable performance, with the added benefit of call sentiment analysis. Some users have noted occasional robotic voice and a clunky workflow builder, but overall, Dialora is a solid choice for those in need of an AI voice agent. Additionally, with a 60-day money-back guarantee, it's worth giving Dialora a try."

AI-powered summary of customer reviews
Showing 1 - 5 of 57 reviews
Avinash.Katta

Verified purchaser

Deals bought: 178Member since: Aug 2014
3 stars
3 stars
Feb 24, 2026

No Indian Numbers to Choose Directly.

Setup is surprisingly fast; you can get started in hours, not weeks.

The AI works well right out of the box without needing tons of training.

One user actually made back their entire purchase cost from just one sale, which is impressive.

The founders are active and respond to issues quickly.

Founder Team
Avi_Dialora

Avi_Dialora

Feb 25, 2026

Thank you!

Working on the Indian numbers. We do have them available but theres a lot of compliance on getting it right now

Helpful?
0
Share
Shopaholic

Verified purchaser

Deals bought: 2825Member since: Dec 2018
2 stars
2 stars
Feb 7, 2026

Credit versus actual minute. Real cost is too much.

We bought 3 of these at max tier 5. It was an exciting moment to see a product that can do what we expected.

However, our excitement was short lived. The support is not bad. They usually replied next day. I am not sure why other users complained about their response time. I had good experience with timely support.

What turned us off is the $1 per minute rate. I am not sure why there is a need to inflate the number to 10 times the actual minute and refer to that number as credits? For example, 10,000 credits and 5,000 credits, etc. The actual cost is .10 per credits, and to to add 10000 credits would give actually 1,000 minutes. This equivalent to $1 per minute.

Why not just list the credits as what they are, as per minute. Inflating the number is just a bit misleading. Any way, I wish them luck at this rate. We ended up refunded all 3 deals. This is really too bad. We really hoped it could work.

Founder Team
Avi_Dialora

Avi_Dialora

Feb 18, 2026

Hi Shopaholic,

Firstly, thank you for the balanced review. I genuinely appreciate that you called out the support experience fairly. That means a lot.

Now let me clear up the math because there seems to be a misunderstanding here.

The Credit System Explained (Very Clearly)

1 minute = 10 credits

If you buy:
• 1,000 credits
• That equals 100 minutes

At Tier 5:
• Overage is $0.10 PER MINUTE...

Helpful?
4
Share
Communicamp

Verified purchaser

Deals bought: 64Member since: Nov 2023
3 stars
3 stars
Jan 24, 2026

we really really wanted to like this service, but not ready for prime time yet

We wanted to use this platform to achieve two objectives:
1) filter incoming calls and pass along to a human only the good ones for our own business
2) Resell the service to our customers as well.
Since AI is the direction the world is heading, it made sense. And the structure of this LTD seemed also fair and sustainable with a purchase of credits that sounded correct.
We purchased a number from Telnyx (we recommend you get it externally since you want to own the number in the future should you decide to part with the service). It took a little bit for KYC verification but then it was perfectly integrated.

Unfortunately, we had a very poor experience with the Dialora AI as it struggled to understand the requests in the Italian language —even basic information like phone numbers and names. Some French Sumolings reported the same.

Overall, despite twiking the instructions for the AI agent, it behaved very differently from what a human would do. And since the AI that responds to a call is supposed to represent the brand or the company, it must be on par with the experience you would like your customers to have.

All cals felt slow, sluggish, and you could tell on the other end there was a robot, not a real person. Especially after changing the initial voice, that we couldn't find anymore, all the others optimized for Italy felt very out of focus.

We didn’t explore the AI capability to understand the technical staff (we would have had to provide much knowledge since there is no way to scan an entire website to get it directly) but we stopped early because all the calls felt fake, and most of all, the AI stopped responding or just hanged there in mute.

We believe the product is a great idea but not ready for prime time yet. We bought lots of deals here on Appsumo that promised great things but only a few actually got better. For that reason, we dediced not to take a risk and we refunded it. Very Sadly because this service is indeed something that could be easily resold to Clients!

Founder Team
Avi_Dialora

Avi_Dialora

Feb 18, 2026

Hi Communicamp,

First, thank you again for giving Dialora a serious try. I completely understand why you made the decision you did at the time — if the AI doesn’t feel brand-safe, it’s not worth the risk.

That said, I do want to clarify something important.

Since your testing, we’ve made significant upgrades:

• Latency is now under ~400ms on average
• Turn-taking and interruption handling have...

Helpful?
1
Share
info3346L

Verified purchaser

Deals bought: 193Member since: Jan 2022
2 stars
2 stars
Jan 8, 2026

Works perfectly in English, but other languages need solid improvements

Good interface, well translated. But, we have tried several calls in German and French. It doesn't understand a word of what is said, which is frustrating. We have used our prompt and also tried with the demo ones (modifying all that was related to the languages). The problem remains. After checking the transcription, we have seen that even if the French/German language is selected, the transcription of what the "client" says is in English BUT the transcription of what the agent says is in French/German. If, during the conversation, the agent thinks that the client speaks English, that could be one of the reasons for the misspellings and other appointment misunderstandings (it didn't understand what "après-demain" (after-tomorrow) means and kept asking for a date; or took the number of people for a name: "3 personnes" (for three), which became "Christine" :-) .

We took the highest tier, and we are not able to resell it because of this solid issue. We keep thinking that the product has good potential and will wait until the last days to request a refund to see if there are improvements, at least if the quality of the European languages is as good as the English version.

Founder Team
Avi_Dialora

Avi_Dialora

Feb 18, 2026

Hi info3346L,

First — thank you for such a detailed and fair review. You clearly tested this seriously, and I respect that.

You are absolutely right about what you observed.

Earlier builds had a language routing issue where:

• The agent voice would speak in French/German
• But the STT (speech-to-text) layer sometimes defaulted to English
• Which caused mis-transcriptions
• Which then broke...

Helpful?
2
Share
James_1

Verified purchaser

Deals bought: 48Member since: Oct 2023
2 stars
2 stars
Dec 18, 2025

Great interface, but scrappy execution

The product has a great interface, appealing features, and good potential, but it is still too far to be even considered in a professional setup.

The robot doesn't understand 2/3 of what is said, cuts me off multiple times, is slow to answer, and has often difficulty speaking, especially in foreign languages such as French.

Any of my clients who would interact with the Dialora robot will have a terrible opinion about me. I can't let that happen. Sad because I believe I have found THE SaaS.

Founder Team
Avi_Dialora

Avi_Dialora

Dec 19, 2025

Hi James, thank you for taking the time to write this. I genuinely appreciate the honesty.

Based on what you’re describing (misunderstanding, interruptions, slow responses, and issues in French), there’s a very high chance this is a prompt and configuration issue, not a core limitation of the platform itself. We see this exact behavior when prompts are too long, too abstract, or not structured...

Helpful?
5
Share