Enhancements -- Prompt library for convenient re-use of teckel AI prompts -- Chat with MCP-tool enabled LLMs (Gemini, ChatGPT, Claude) -- Asynchronous response mode (with in-app notifications) for long-running teckel AI jobs (for MCP-enabled LLMs) -- IPFS backend-storage for persisting responses from teckel AI LLMs Plus minor UI improvements and big-fixes

Release 1.1.6: Notifications and address book feature introduced to teckel talk. You can now be notified whenever someone in your address book -- or in any of your private chat groups -- joins the chat. Improved private chat group management, making it easy to add anyone to your private group(s) or address book directly from the chat interface. Release 1.1.5: hotfix for version 1.1.4 to correct an issue with teckel AI access to Gemini Release 1.1.4: Teckel AI now supports simultaneous chat with Gemini, ChatGPT, Perplexity, DeepSeek, and Claude (previously just Gemini and ChatGPT). Teckel AI now supports selection of common model variants within each supported LLM family. For example, for ChatGPT, the user can choose between ChatGPT 4.1 & 5, or -the respective mini and -nano variants. Teckel AI allows the user to adjust the selected model parameters, such as "temperature", where relevant. Teckel AI incorporates a convenient mechanism for generating "system prompts" to provide general instructions to the LLMs e.g., "Be short and concise". Introducing the "CrossListen" feature to Teckel AI whereby the user can "point at" a subset of LLM participants within a conversation, and direct only them to respond to the upcoming prompt. The others will "listen quietly", in preparation for providing their feedback afterwards, on request. Improvements to the "CrossTalk" feature (introduced in the previous update) which facilitates head-to-head interaction between two or more participating LLMs, contributing to the overall conversation. Together, "CrossListen" and "CrossTalk" facilitate natural, flowing conversation between the user and multiple LLMs simultaneously.

Release 1.1.5: hotfix for version 1.1.4 to correct an issue with teckel AI access to Gemini Release 1.1.4: Teckel AI now supports simultaneous chat with Gemini, ChatGPT, Perplexity, DeepSeek, and Claude (previously just Gemini and ChatGPT). Teckel AI now supports selection of common model variants within each supported LLM family. For example, for ChatGPT, the user can choose between ChatGPT 4.1 & 5, or -the respective mini and -nano variants. Teckel AI allows the user to adjust the selected model parameters, such as "temperature", where relevant. Teckel AI incorporates a convenient mechanism for generating "system prompts" to provide general instructions to the LLMs e.g., "Be short and concise". Introducing the "CrossListen" feature to Teckel AI whereby the user can "point at" a subset of LLM participants within a conversation, and direct only them to respond to the upcoming prompt. The others will "listen quietly", in preparation for providing their feedback afterwards, on request. Improvements to the "CrossTalk" feature (introduced in the previous update) which facilitates head-to-head interaction between two or more participating LLMs, contributing to the overall conversation. Together, "CrossListen" and "CrossTalk" facilitate natural, flowing conversation between the user and multiple LLMs simultaneously. Minor UI improvements Minor bug fixes

Teckel AI now supports simultaneous chat with Gemini, ChatGPT, Perplexity, DeepSeek, and Claude (previously just Gemini and ChatGPT). Teckel AI now supports selection of common model variants within each supported LLM family. For example, for ChatGPT, the user can choose between ChatGPT 4.1 & 5, or -the respective mini and -nano variants. Teckel AI allows the user to adjust the selected model parameters, such as "temperature", where relevant. Teckel AI incorporates a convenient mechanism for generating "system prompts" to provide general instructions to the LLMs e.g., "Be short and concise". Introducing the "CrossListen" feature to Teckel AI whereby the user can "point at" a subset of LLM participants within a conversation, and direct only them to respond to the upcoming prompt. The others will "listen quietly", in preparation for providing their feedback afterwards, on request. Improvements to the "CrossTalk" feature (introduced in the previous update) which facilitates head-to-head interaction between two or more participating LLMs, contributing to the overall conversation. Together, "CrossListen" and "CrossTalk" facilitate natural, flowing conversation between the user and multiple LLMs simultaneously. Minor UI improvements Minor bug fixes