New features and improvements in SiteSpeakAI, newest first.
Get access to the latest GPT models from OpenAI with SiteSpeakAI's AI Agents. You can now update your AI Agents to use GPT-4.1 (or GPT-4.1 mini and GPT-4.1 nano) to improve their knowledge and capabilities even more. Read more about the latest OpenAI models here.
Anthropic released the latest and most intelligent version of their Claude model yesterday (24 February 2025). The new model, Claude 3.7 Sonnet, can produce near-instant responses which means faster replies to your users questions and queries. To get started using the latest Anthropic model, go to Settings -> Basic for your chatbot, and choose the anthropic/claude-3-7-sonnet model from the dropdown.
You can now add your custom AI agent to your Discord server 🎉. You can also add channels from your Discord server as training sources to your chatbot to enhance it's knowledge and allow it to respond to member queries using previously answered questions from your Discord server. This means you can have a single AI agent operate as a customer support chatbot on your website, Telegram, Slack and Discord, using knowledge from all your sources to respond to queries.
You can now use the most recent Google Gemini 2.0 Flash model in your customer support chatbots. To try Gemini 2.0 Flash today in your chatbot, go to Advanced -> Settings -> AI Model and select "google/gemini-2.0-flash" from the dropdown menu. Click "Update" and your chatbot will now be using this new model for all queries.
You can now use the most recent OpenAI reasoning model in your customer support chatbots. o3-mini can outperform o1 in reasoning tasks, and has lower latency. This means better and faster responses to your user queries. To try o3-mini today in your chatbot, go to Advanced -> Settings -> AI Model and select "openai/o3-mini" from the dropdown menu. Click "Update" and your chatbot will now be using this new model for all queries.
You can now try 4 new models for your customer support chatbots. We've added support for Meta Llama 3.1 405B, 3.1 8B and 3.3 70B as well as Cohere Command R.
You can now easily generate a free llms.txt for your website to provide information to help LLMs use your website at inference time. Read more about the proposal for a llms.txt standard here.
The newly launched (last 24 hours) Grok Beta model from xAI is now available to use in SiteSpeakAI. To enable, go to Advanced -> Settings -> AI Model and select the new grok-beta option.
You can now update the default message shown when you enable escalations for your chatbot or when you've added the "Escalate to Human" smart prompt for your chatbot. To set your custom message, go to Advanced -> Settings -> Localization and update the message under the Escalations section.
Join our official Customer Support Telegram group to get all your questions about SiteSpeakAI and creating and installing your own custom GPT chatbot or assistant answered. Join here: Telegram Link
Join 1,000+ businesses across SaaS, e‑commerce, and agencies automating their customer service and other tasks with a custom-trained AI agent.
No credit card required