Gali
Gali is an AI chatbot tool specializing in generating intelligent and personalized customer interactions on websites. It leverages AI GPT t...
Last verified:
What is Gali?
Gali AI is a platform that allows users to create custom ChatGPT assistants trained on their own data, answering questions in real time and available 24/7. The platform is designed to handle standard questions automatically, reducing the workload for teams and allowing them to focus on more advanced tasks that require human intervention.
Key features include no-code setup with a 10-minute quick deployment process, web crawling to train chatbots from website URLs, multiple file upload for documents, and complete Zapier integration for seamless tech stack connectivity. Gali supports GPT-4 and GPT-3.5 engine models, offers chat analytics to learn insights from interactions, and automatically responds in all languages supported by OpenAI including English, French, Italian, Spanish, German, Portuguese, and more.
The platform is ideal for businesses looking to integrate AI chatbots into their websites and apps for customer support, lead generation, sales assistance, and internal knowledge Base access. Use cases include Salesbot for transforming website visitors into leads, Customer Support for answering inquiries, Wiki & docs for internal documentation access, and fully-tailored custom prompts for specific requirements.
Gali offers seamless integration through HTML snippets that can be embedded on any website, API access for advanced usage, and integrations with WhatsApp, Slack, and Freshworks. The platform includes features like escalating chats to human agents by email, capturing lead contact details directly in chat, and monitoring answers to improve training quality.
Gali pricing
Pricing model: Freemium
Gali AI offers four pricing tiers with monthly and yearly options. Test plan is free with 1 chatbot, 30 message credits/month, 400k max source characters, 30-day chat history. Indie plan is $19/mo ($190/y) with 3 chatbots, 2,000 message credits/month, 8M max source characters, 90-day history, GPT-4, escalate to human, capture leads. Startup plan is $99/mo ($990/y) with 6 chatbots, 10,000 message credits/month, 10M max source characters, 365-day history, premium support, APIs, integrations (Zapier, WhatsApp, Slack, Freshworks), function calling. Scaleup plan is $249/mo ($2,490/y) with 10 chatbots, 25,000 message credits/month, 15M max source characters, 365-day history, guided onboarding. Pay-as-you-go for extra credits: $0.015 (Indie), $0.012 (Startup), $0.01 (Scaleup). Monthly plans renew every 30 days, yearly every 360 days.
Gali pros
- No coding required for setup
- 10-minute quick setup process
- GPT-4 available at accessible rates
- Complete Zapier integration
- 24/7 availability and infinitely scalable
- Trained on your own custom data
- Multi-language support in all OpenAI languages
- Web crawling from website URLs
- Multiple file upload capability
- Chat analytics for insights
- Escalate to human agent by email
- Capture leads directly in chat
- API access for advanced usage
- Integrations with WhatsApp, Slack, Freshworks
- Function calling to connect endpoints
- Suggested prompts for inspiration
- Customizable chatbot appearance
- Compatible with Custom HTML and any platform
- Pay-as-you-go for extra message credits
- 365-day chat history on higher plans
Gali cons
- GPT-3.5 can hallucinate more than GPT-4
- Prompt engineering requires trial and error
- Extended context uses more credits (2-10 per message)
- Temperature must be kept low (0-0.2) for consistency
- Maximum source characters limited per chatbot plan
- Pay-as-you-go feature still coming soon
- GPT-4 is 6x more expensive than GPT-3.5
- Free tier only has 30 message credits/month
- VAT not included in pricing
- Slack and Microsoft Teams integrations still planned
Frequently asked questions about Gali
What is Gali AI?
Gali AI is the platform to create custom ChatGPT assistants trained on your data, answering in real time, and available 24/7. It's like chatting with ChatGPT, but answering based on the data you provide from your website, documentation, FAQs, and other sources.
How long does it take to set up Gali?
You can create your own chatbot in just 10 minutes. The setup process includes: Sign-up (1 min), Customize appearance (3 min), Add sources and train (2 min), and Add to your website (4 min). You can test it within approximately 5 minutes.
Do I need coding skills to use Gali?
No coding is required. Gali offers no-code, quick setup where you just define some parameters and the platform takes care of the rest. It's extremely easy to add on any website using HTML snippets from the Export section.
Which AI engine models does Gali support?
Gali offers GPT-3.5 and GPT-4 engine models. GPT-3.5 is cheaper but can hallucinate more and sometimes takes liberties with following prompts. GPT-4 provides higher quality output, especially for complex questions, follows prompts much better, and is very unlikely to hallucinate. GPT-4 is 6x more expensive on their pricing.
What sources can I use to train my Gali chatbot?
You can train Gali chatbots using website URLs (via web crawling), multiple document files, FAQs, documentation, and custom documents. When using website sources, you just provide the link and Gali handles selecting relevant pages, fetching data, and adding them as training sources.
Does Gali support multiple languages?
Yes, Gali bots powered by OpenAI GPT are capable of training and responding in multiple languages. There is no need to preset the response language as bots automatically respond in the language of the user's input. Supported languages include US English, UK English, French, Italian, Spanish, German, Portuguese, Brazilian Portuguese, and many more.
How do I add Gali to my website?
Once configured on the dashboard, navigate to the Export section to obtain the HTML snippet. This code can be incorporated into your website. Options include placing an icon in the lower right corner or using an iframe section. Gali is compatible with Custom HTML and any platform including Webflow and WordPress.
What is the MECE principle for documentation?
MECE stands for Mutually Exclusive, Collectively Exhaustive. Exhaustive means including all relevant information across documents to cover topics fully. Exclusive means avoiding overlapping information to prevent confusion. This principle ensures as many questions as possible are answered with very precise content.
What temperature setting should I use for Gali?
Gali suggests keeping temperature low (0 to 0.2) to avoid too much randomness in answers, especially for customer support or when you don't want weird surprises. A temperature of 0 means the model selects the word with highest probability, while higher temperatures (0.8-1) lead to more variation and creativity.
Can I integrate Gali with my tech stack?
Yes, Gali offers complete Zapier integration to instantly push new messages to your tech stack and trigger growth campaigns when capturing leads. It also includes integrations with WhatsApp, Slack, and Freshworks on Startup and Scaleup plans. APIs are available for advanced usage, and function calling lets you connect your endpoints.