re:tune
re:tune is a comprehensive platform designed for building AI applications without the need for coding. The platform facilitates the creation of custom AI applic...
Last checked:
What is re:tune?
re:tune is a no-code AI app building platform that serves as the missing frontend for AI models from OpenAI, Anthropic, and other providers. The platform enables users to create, customize, and integrate custom AI chatbots similar to ChatGPT without writing any code. It provides a simple and intuitive UI that removes the complexity of building AI chatbots, making it accessible for personal users, startups, agencies, and businesses looking to transform their operations with AI.
Key features include the ability to build chatbots for any use case from customer support to sales and lead generation, train AI models with any data source (website content, PDFs, text documents, YouTube videos), and integrate chatbots across multiple platforms. The platform offers website crawling for training, infinite memory in chatbots to remember all previous interactions, auto-prompt features to generate detailed prompts from short descriptions, custom widget customization, REST API access, and the ability to create custom fine-tuned models based on user datasets. Users can publish chatbots as website widgets (pop-up or iframe), share via link, or integrate with WhatsApp, SMS (Twilio), and soon Messenger, Telegram, Instagram, Slack, and Zapier.
re:tune is designed for individuals building personal assistants, startups and small businesses needing customer support chatbots, agencies managing multiple client chatbots, and teams with custom security and performance needs. The platform supports 10+ AI models from OpenAI, Anthropic, Google, Meta, and more including GPT-4, GPT-3.5, and Claude. Users bring their own OpenAI or Claude API keys, allowing unlimited users and messages without artificial constraints.
re:tune pricing
Pricing model: Paid
re:tune offers a free Hobby tier and four paid plans. Hobby (Free): 10 personal chatbots, publish on website, 1000 messages per chatbot, 100 users per chatbot, 5 file uploads/month. Launch ($499/year): Everything in Hobby plus unlimited personal chatbots, unlimited messages per chatbot, unlimited users per chatbot, website crawling training, 50 file uploads/month, infinite memory, customize chat widget, early access to new features, unlimited usage history. Grow ($999/year): Everything in Launch plus 5 published chatbots, 5 workspaces, 500 file uploads/month. Scale ($1999/year): Everything in Grow plus 10 published chatbots, 10 workspaces, unlimited file uploads/month, remove branding, sync with website crawling (coming soon). Max ($3999/year or $399/month): Everything in Scale plus unlimited published chatbots, unlimited workspaces, custom domain and whitelabeling, custom trained fine-tuned models, custom integrations with any application. All paid plans include 30-day money-back guarantee and free trial option. Users must bring their own OpenAI API key (pay-as-you-go, $0.1 per million tokens for embedding).
re:tune pros
- No-code interface requires no programming knowledge
- Truly unlimited personal chatbots with own API keys
- Unlimited messages and users without artificial constraints
- Infinite memory remembers all previous chat interactions
- Train with any data source: websites, PDFs, text, YouTube
- Website crawling feature for automatic training data collection
- Auto-prompt generates detailed prompts from short descriptions
- 10+ AI models available including GPT-4, GPT-3.5, Claude
- Custom fine-tuned model creation from user datasets
- Multiple publishing options: widget, iframe, shareable link
- WhatsApp and SMS (Twilio) integration already available
- REST API for creating chatbots, threads, and continuing conversations
- Custom widget customization (name, icon, color, position, greetings)
- Custom domain and white-labeled branding available on Max plan
- 30-day money-back guarantee on all paid plans
- Early access to new features on Grow and Launch plans
- Priority feature requests for paid plan users
- Multiple workspaces for separate client projects
re:tune cons
- Requires users to bring their own OpenAI/Claude API key
- Need active OpenAI account with payment method added
- XML sitemap support still in development (HTML only currently)
- No user authentication feature during chat yet
- Cannot disable manual input for predefined button options only
- Dashboard still in English only (translation coming in ~2 months)
- Not SOC2 compliant yet (planned in couple months)
- No human agent/takeover support yet (coming in 3-4 weeks)
- Cannot provide information source references yet (on roadmap)
- No formal file size limit but uploads failing if over 1 minute
- Microsoft Teams integration not planned yet
- Webhooks still on roadmap (not available yet)
- Upvote/downvote answer training feature still in progress
- Lead capture fields feature still on roadmap
- Data stored only in US (EU region coming soon)
Frequently asked questions about re:tune
Do I need to purchase a model plan to create a bot?
Not really, you need to bring your own key to use re:tune. Hence you will need an active OpenAI account with a payment method added to it. You can sign up at platform.openai.com (not chatgpt) and add a payment method. You don't need a subscription as OpenAI is pay-as-you-go.
What is the difference between a personal chatbot and a published chatbot?
A published chatbot can be embedded in a website as a pop-up or iframe widget. A personal chatbot is only accessible from the dashboard and is meant for personal use only. Personal chatbots are internal, while published chatbots are external and can be added to websites. With T4 (Max) plan, you can add your bot to unlimited websites.
Can I set limits on tokens or messages for my chatbots?
Yes, go to Chat > Settings > Limit to set message limit for each thread. Currently re:tune supports setting limits for the number of threads and the number of messages per thread. You can also set limits on output length, numbers of messages a user can send every day/week/month, and numbers of new threads that can be created.
How can I train my bot with all links from my sitemap?
re:tune currently supports HTML sitemaps with links in <a> tags. XML sitemap support is in progress and coming soon. For the time being, you can either use a HTML sitemap or scrape up to 200 URLs at once by pasting them in the dashboard.
What integrations are available?
Right now, re:tune has WhatsApp and SMS (with Twilio) integration for training. Messenger, Telegram, Instagram, Slack, and Zapier are coming very soon. There is also YouTube video integration for training. The team is working on Zapier integration specifically, creating triggers and actions that will make it easier to integrate with other automation tools like n8n and make.com shortly after.
Can I assign different API keys to different chatbots?
You can't directly assign an API key to a bot, but you can create separate workspaces and assign different API keys to each workspace. So if you have chatbots in separate workspaces, you can use different API keys. With T4 (Max) plan, you can use unlimited API keys across unlimited workspaces.
What is custom domain and how many can I use?
Usually when you embed a chatbot or share any thread, it will have a retune.so link (e.g., https://retune.so/...). With the custom domain feature, you can have your own domain to host and share the chatbots (e.g., https://yourdomain.com/...). Custom domains are per workspace, so you can have multiple. With Tier 4 (Max), you can configure separate custom domains per workspace.
Are there limits on users/messages that can parallelly converse with our bot?
No, there's no limit on that! By allowing you to bring your own keys, re:tune doesn't put any artificial restrictions on the number of users or messages. You can have unlimited users and messages because since you are bringing your own key, the platform doesn't need to put artificial constraints on usage.
How do I prevent misuse of my API key with endless questions?
You can set limits on: length of output, numbers of messages a user can send every day/week/month, and numbers of new threads that can be created. These settings are found in Chat > Settings > Limits. re:tune is working on even more ways to control usage and prevent misuse, like rate-limiting based on IP address and token count.
Can I use re:tune for my chatbot agency and sell it to other businesses?
Yes, you can definitely sell the chatbots since they can be published on any website. You can even use your own custom domain or subdomain in the embed code. The Max plan includes white-labeled branding for chatbots and custom trained fine-tuned models, which is ideal for agencies. You can create workspaces for separate client projects and potentially invite clients to view their chatbot and conversations (Seats feature coming soon).