Assistants are the quickest way to get started in building LLM-Powered chat applications with Griptape ☁️ Cloud. If you want see exactly how quick, sign up for a free Griptape Cloud account here ▶︎ and then check out our new guided assistant walkthrough ▶︎ This sample uses up-to-the-minute data about Super Bowl 🏈 LIX, to create an Assistant that can answer questions about the game and all the happenings around it. We provide the large language model that powers Griptape Cloud Assistants, so in just a few clicks you can create your own Assistant and see how easy it is to combine up-to-date data with a large language model in Griptape Cloud.

We're pleased announce availability of Griptape 🛹 Framework version 1.3. There are a lot of new features in this release that have been inspired by feedback from Griptape customers and members of the open source community including: * Support for OpenAI's o1/o1-mini and o3-mini Reasoning 🧠 Models. We also already have support for DeepSeek through our OpenAI compatible driver, but perhaps that's been hyped enough already, so we are burying that information at the end of this bullet point 😂 * Support for Audio Input and Output 🔊 in the OpenAIChatPromptDriver * A new GriptapeCloudStructure util that makes it simpler to configure Structures running in Griptape ☁️ Cloud to use the services there, such as the EventBus 🚌 and Observability 🔎 features. * A new LocalRerankDriver that calculates RAG query and search result similarity without the need to hit Cohere's Rerank API endpoint. Of course, if you still want to use Cohere's Rerank model, you can. This is an additional driver. * Further Structured Output improvements with support for pydantic.BaseModels. When using pydantic.BaseModels, Structured Output will return instances of the models that you define, so they support type checking and hints. We recommend you use them for your Structured Output needs 😀 Improvements to the DateTimeTool ⏰ for calculating time deltas and differences between dates. If you want to learn more, please check out our blog that describes the most significant new features in a more detail. You can find that here:

RT @_andrewfrench: I gave a @griptapeai assistant a schema and a some rules, then wired up the response to a simple Chart.JS frontend to qu…

Sample structures to get started with both Griptape and Griptape Cloud:

RT @EDBPostgres: AI moves too fast for rigid systems & vendor lockin. 30% of #GenAI projects fail because they lack a foundation built to l…

RT @Amaru_Zeas_3D: Life Hunter World is coming. This image has a combination of #3d, #cgi and #ai.@Substance3D for assets, @Free_EpicGames…

RT @austinaifilm: @aasanch of the Austin AI Film Festival sits down with former Amazon Creative @Amaru_Zeas_3D who is now Creative Director…

Structured Output is a critical feature for developers building applications with large language models. How can you reliably work with the output from LLMs or build logic based on that output if the structure is unpredictable? The answer is, of course, that you can’t. Version 1.2 of the Griptape Framework adds Structured Output support for all of our supported models, including models from @OpenAI , @AnthropicAI , @ollama , @awscloud and more. Learn how to use this Structured Output in your LLM-powered applications in our latest blog: #LLM #griptapeai #DeveloperTools

We're pleased to announce that Griptape is now available via the Amazon Web Services (AWS) Marketplace. Griptape's end-to-end AI application stack enables developers to efficiently build, deploy, and manage secure LLM-powered AI applications with predictable output. It gives developers everything they need, from the development framework to the execution runtime. AWS Marketplace allows you to use Griptape Cloud with a simple monthly license fee, billed through your existing AWS account. You can find Griptape on the AWS Marketplace here: or you can simply search for 'Griptape' in the AWS Marketplace search.

Want to build AI applications five times faster without getting bogged down in prompt engineering and ending up with unreliable results? The Griptape Framework is your answer! 🏇 5x faster development with predictable, programmable Python interface 🛡️ Enhanced security with Off-Prompt™ Technology 🎯 More reliable outputs without prompt engineering ‘expertise’ 🔧 Complete toolkit for Agents, RAG, and AI pipelines ⚡️ Simple setup with pip/Poetry installation In just minutes, you can build AI agents, ETL (Extract, Transform, Load) pipelines, and RAG implementations that actually work. Getting started is as easy as `pip install ‘griptape[all]’`. Learn how to get started with the Griptape Framework and join the growing community of developers building the next generation of AI applications!  #AI #Development #Python #LLM #GriptapeAI

Love the series of videos @fahdmirza has created on @griptapeai - Griptape - AI Agents With Reasoning, Tools, and Memory - RAG with Griptape - Modular AI Agents Using RAG - Hands-on Demo - Use AI Agents for Transcription with Griptape - Step-by-Step Tutorial

Happy New Year! Griptape 1.1 has been released, now available on PyPi! All features/changes in this release are 100% backwards compatible with 1.0. Read more here: Conversation Memory Strategy: Control how Conversation Memory entries are created using the new Structure.conversation_memory_strategy. Task Conversation Memory: Use Prompt Tasks with Conversation Memory without needing a Structure. Branch Task: A new Task type that allows you to create branches in your Workflows. Eval Engine: A new Engine that allows you to evaluate the quality of generated text. Structure Run Stream: A new way to stream events from Structure runs. Prompt Task Tools: Prompt Task now supports Tools, removing the need for a separate Toolkit Task. Talk To Document: AnthropicPromptDriver and AmazonBedrockPromptDriver now support GenericMessageContent, allowing you to talk to a document.

RT @griptapeai: Discover the power of Griptape Cloud’s core features in this hands-on demo! We’ll explore essential components like Data Co…

RT @gm_mertd: Introducing Associates 🫡: Your drop in relentless go-to-market partner. The problem we are solving is exclusive to those wh…

Create a Custom AI Chat Assistant in Slack with Griptape in Under 5 Minutes! This tutorial showcases how to leverage Griptape Cloud to build an assistant that can access knowledge bases, respond intelligently to queries, and provide actionable insights. #GriptapeCloud #aichatbots #ai #techdemo #aiagents #slackbot

RT @jturow: AI agents feel stuck in first gear—few products at scale today. But innovation below the surface (e.g. Stripe agent payment SDK…