1st. I am trying to sent up my chat to reach out to my server using a webhook to look for info for questions on data that I have stored there. Through detailed logs & trouble shooting I see the request leave the Tiny Talk bot hit my server, my server responds with the correct info but the info is never displayed in the Tiny Talk bot chat window. What format does the return info need to be in to show in the embed Iframe window? Nothing seems to work.
2nd. It states that the chat bots have API keys for paid accounts, I have a paid account but cannot find the location of the bots API keys. Thanks
Webhooks are inherently one way only. They are designed to send data from one system to another when a specific event occurs. In our case when a certain event happens in your chatbot, then we send a request to an endpoint you provided, responses are not listened or acted upon.
I'm not sure what made you think this was possible, please let us know if we failed to communicate something around this well enough.
You can create an API Key for your Tiny Talk Chatbot under the Integrations page, using Tiny Talk section.
I think it would be best if you reach out by email at support@tinytalk.ai and feel free to describe your use case and intent, so we can better understand your goals, what you're trying to achieve and provide you the best support we can.
20M characters: across all chatbots, correct. 20,000 messages per month: across all chatbots, correct.
All our AppSumo plans require you to bring your OpenAI API Key, we do not offer any free AI usage with paid accounts. The limits are in place to allow fair usage of our platform and to account for the compute/bandwidth and infrastructure that processes those messages. Our free plan has 25 free messages per month for you to test the platform before committing.
WhatsApp integration does not require an API key, you'd login using your Facebook/Meta account and follow 4-5 step onboarding flow to complete integration. This is called "Embedded Signup", see the following page for further information: https://developers.facebook.com/docs/whatsapp/embedded-signup/
WhatsApp Integration: I am not sure if I got your question right, but when the phone number integrated with Tiny Talk receives/sent a message, it is still counted against your monthly message limit.
Let me know if you have further questions, happy to help!
Q: Title: Questions about Multimedia and Audio Capabilities in Tiny Talk
1. Is it possible to send and receive multimedia messages (images, videos, audios) through the Tiny Talk chatbot? 2. Can the Tiny Talk chatbot process and respond to audio messages sent by users? 3. Can a human agent integrated in the “Human Takeover” feature respond with audio messages? 4. Does Tiny Talk offer any functionality to convert audio messages into text so the chatbot can interpret them? 5. What multimedia file formats are supported by the platform in chat conversations? 6. Are there any limits or restrictions on sending multimedia files during conversations? 7. Can the chatbot be configured to send pre-recorded audio messages as automatic responses?
At the moment no, but this is something we want to introduce eventually, it is part of our roadmap but we have large features to deliver first before we get to that. This is mostly valid for the remaining questions.
Let me know if there is anything I can help with further.
Thank you for reaching out. You can enable "Human Takeover" for your chatbot and users will be offered the option to request human takeover when they feel like it. When this happens, your chatbot is still in charge until you explicitly take over the conversation by sending a message to the conversation/reply to the user requesting it.
Does this address your question? Let me know if you meant something slightly different or have further questions, I'd be happy to help!
I bought the first-tier deal recently and tried setting up my first bot. I added 6 documents and a total of about 55K characters. I clicked to train the bot based on these data but for over 4 hours, the training isn't done yet. How long does it take? And also, does it mean I can't have any interactions with the bot unless the training is completed?
Thank you for reaching out. We've replied to your email, but I'll provide the answer here as well for visibility:
We have a queue for training, when you submit your resources they are queued and if there were other accounts who have queued their resources before yours then your resources would wait for those to complete first.
Generally you don’t wait for long, it happens almost right away, but sometimes the queue might be busy due to a long running job and you might need to wait longer than usual.
We don’t notify you when training is complete.
As for responses, we don’t wait for any training to complete, your bot would always respond if there is nothing prevents it from responding like lack of an API key.
If your bot isn’t responding at all, it is most likely due to a missing API key or lack of credits in your OpenAI account.
Hope this helps and let us know by email if you continue having an issue, I'd be happy to take a closer look.
Q: Webhook functionality & API
Hello,
1st. I am trying to sent up my chat to reach out to my server using a webhook to look for info for questions on data that I have stored there. Through detailed logs & trouble shooting I see the request leave the Tiny Talk bot hit my server, my server responds with the correct info but the info is never displayed in the Tiny Talk bot chat window. What format does the return info need to be in to show in the embed Iframe window? Nothing seems to work.
2nd. It states that the chat bots have API keys for paid accounts, I have a paid account but cannot find the location of the bots API keys.
Thanks
tinytalk
Jun 25, 2025A: Hello!
Thank you for reaching out.
Webhooks are inherently one way only. They are designed to send data from one system to another when a specific event occurs. In our case when a certain event happens in your chatbot, then we send a request to an endpoint you provided, responses are not listened or acted upon.
I'm not sure what made you think this was possible, please let us know if we failed to communicate something around this well enough.
You can create an API Key for your Tiny Talk Chatbot under the Integrations page, using Tiny Talk section.
I think it would be best if you reach out by email at support@tinytalk.ai and feel free to describe your use case and intent, so we can better understand your goals, what you're trying to achieve and provide you the best support we can.
All the best,
Oscar
Share Tiny Talk
Verified purchaser
Thank you this has been very helpful.
Q: Clarification :)
Hello!
I have a few questions please. For example, based on tier 2:
20M characters (across all chatbots right?)
20,000 messages per month (across all chatbots right?)
GPT 3.5 (Bring your own key) (if I used below 20k messages, do I still need an OpenAI key? Or it works up to 20k with your own key?)
WhatsApp integration (I would also need a Whatsapp API key and all messaging is charged outside TinyTalk? Or it's also part of the 20k messages?)
Thanks!!
tinytalk
Jun 23, 2025A: Hi Bernardo!
Thank you for reaching out.
20M characters: across all chatbots, correct.
20,000 messages per month: across all chatbots, correct.
All our AppSumo plans require you to bring your OpenAI API Key, we do not offer any free AI usage with paid accounts. The limits are in place to allow fair usage of our platform and to account for the compute/bandwidth and infrastructure that processes those messages.
Our free plan has 25 free messages per month for you to test the platform before committing.
WhatsApp integration does not require an API key, you'd login using your Facebook/Meta account and follow 4-5 step onboarding flow to complete integration. This is called "Embedded Signup", see the following page for further information:
https://developers.facebook.com/docs/whatsapp/embedded-signup/
WhatsApp Integration: I am not sure if I got your question right, but when the phone number integrated with Tiny Talk receives/sent a message, it is still counted against your monthly message limit.
Let me know if you have further questions, happy to help!
All the best,
Oscar
Share Tiny Talk
Q: Title: Questions about Multimedia and Audio Capabilities in Tiny Talk
1. Is it possible to send and receive multimedia messages (images, videos, audios) through the Tiny Talk chatbot?
2. Can the Tiny Talk chatbot process and respond to audio messages sent by users?
3. Can a human agent integrated in the “Human Takeover” feature respond with audio messages?
4. Does Tiny Talk offer any functionality to convert audio messages into text so the chatbot can interpret them?
5. What multimedia file formats are supported by the platform in chat conversations?
6. Are there any limits or restrictions on sending multimedia files during conversations?
7. Can the chatbot be configured to send pre-recorded audio messages as automatic responses?
tinytalk
Edited Jun 23, 2025A: Hello Juan,
Thank you for reaching out!
At the moment no, but this is something we want to introduce eventually, it is part of our roadmap but we have large features to deliver first before we get to that. This is mostly valid for the remaining questions.
Let me know if there is anything I can help with further.
All the best,
Oscar
Share Tiny Talk
Q: Does this have a live chat or human takeover feature?
Does Tiny Talk offer the option for human takeover/live chat functionality or is it strictly AI chatbot only?
tinytalk
Jun 20, 2025A: Hello!
Thank you for reaching out.
You can enable "Human Takeover" for your chatbot and users will be offered the option to request human takeover when they feel like it. When this happens, your chatbot is still in charge until you explicitly take over the conversation by sending a message to the conversation/reply to the user requesting it.
Does this address your question? Let me know if you meant something slightly different or have further questions, I'd be happy to help!
All the best,
Oscar
Share Tiny Talk
Q: Training Not Running
I bought the first-tier deal recently and tried setting up my first bot. I added 6 documents and a total of about 55K characters. I clicked to train the bot based on these data but for over 4 hours, the training isn't done yet. How long does it take? And also, does it mean I can't have any interactions with the bot unless the training is completed?
tinytalk
Jun 20, 2025A: Hello,
Thank you for reaching out. We've replied to your email, but I'll provide the answer here as well for visibility:
We have a queue for training, when you submit your resources they are queued and if there were other accounts who have queued their resources before yours then your resources would wait for those to complete first.
Generally you don’t wait for long, it happens almost right away, but sometimes the queue might be busy due to a long running job and you might need to wait longer than usual.
We don’t notify you when training is complete.
As for responses, we don’t wait for any training to complete, your bot would always respond if there is nothing prevents it from responding like lack of an API key.
If your bot isn’t responding at all, it is most likely due to a missing API key or lack of credits in your OpenAI account.
Hope this helps and let us know by email if you continue having an issue, I'd be happy to take a closer look.
All the best,
Oscar
Share Tiny Talk