Groq News | Sep 06, 2024

Brought to you by RivalSense - an AI tool for monitoring any company.

RivalSense tracks the most important product launches, fundraising news, partnerships, hiring activities, pricing changes, tech news, vendors, corporate filings, media mentions, and other developments of companies you're following 💡

---

Company Overview

Groq is the creator of the LPU™ AI Inference Technology for fast, affordable, and energy efficient AI. Headquartered in Silicon Valley, Groq provides cloud and on-prem solutions at scale for AI applications. The LPU and related systems are designed, fabricated, and assembled in North America.

Latest Updates

• Gui Bibeau announced a new tool using Groq's technology to convert videos into blog posts in under a minute.
• Groq and Llama 3.1 built a complete API structure in VSCode using CodeGPT's CodeBuilder feature.
• Groq announced the availability of the e2b_dev Code Interpreter in their Cookbook for building LLM apps requiring reasoning.
• Groq is delivering a multi-modal endpoint hosting LLaVA-1.5 7B, which is reportedly 4x faster than GPT-4 based on benchmarking.
• Groq has introduced LLaVA v1.5 7B, a new visual model available on its Developer Console, making GroqCloud multimodal with support for image, audio, and text.
• Oliver Sommer announced that GroqCloud now supports multimodal AI, combining visual, auditory, and textual inputs.
• Groq has launched vision models, including text-to-image capabilities.
• Groq is hosting models for an e-commerce product description generation application built using llava-v1.5-7b Llama-70b by AIatMeta.
• Groq now supports image interaction at high speeds, alongside audio and text.
• Vision models are now available on Groq's playground and API.
• NavamAI now supports models from Groq among other providers.
• Mark Heaps announced a collaboration between Groq, Gradio, and Hugging Face for building voice assistants, offering Groq swag for cool projects.
• Julien GROS has joined Groq as a Private Investor in France, previously working at OpenAI in the same position.
• Groq models, including whisper-large-v3 for audio transcription, were integrated into Roam Research's Live AI Assistant.
• Groq announced a new feature for testing logs on GroqCloud.
• Meta has partnered with Groq to help developers unlock the full potential of its AI models.
• The Llama3 Remote AI Model Communication Plugin for Unreal Engine connects to the Groq API server, featuring advanced interaction capabilities and integrated ElevenLabs API for voice output.
• Martin Bowling announced a new AI project called Speculative Retrieval-Augmented Generation (RAG) that uses Groq's fast inference capabilities.
• Mark Heaps previewed using LottieFiles motion copilot powered by Groq's LLM integration.
• Hossein (Unclecode) added YouTube support to Hermes, a video transcriber using Groq's distil-whisper model.
• Groq demonstrated a system combining Llama 3, Gemma 2, and Mistral models to generate and edit MLB game summaries.
• Groq launched their Distil-Whisper endpoint, the fastest Speech to Text endpoint with a Speed Factor of 240x and a cost of $0.33 per 1000 minutes of audio.
• Groq released Distil-Whisper: distil-large-v3 on GroqCloud, available on their website, console, and via API.
• Groq launched ScribeWizard, an audio-to-notes tool created by intern Benjamin Klieger, allowing users to choose their outline and content generation models.
• Groq transitioned its 1st gen 14nm LPU chip to a 4nm LPU, achieving a 3x improvement even without optimization.
• Groq's technology is now integrated into Vectara's vectara-agentic platform for building Agentic RAG applications.
• Groq's Open WebUI now supports integration with OpenAI-compatible APIs and various platforms like GroqCloud and Mistral for enhanced LLM inference.
• Groq launched an automatic summarization feature for videos and audios, powered by Llama from AI at Meta.
• Pastel Network's Inference protocol integrates major AI models including Groq, Mythral, OpenAI, Swiss Army Llama, and Anthropic.
• Groq released the Llama Guard 3 model for content moderation, which can be used via their API for screening messages against 14 harmful categories.
• Tom Terado introduced DocDoctor, an AI assistant using Groq's llama 3 to produce simple ReadME content from example code.
• Groq completed a challenge using their API inference and deployment with Vercel.
• Groq offers support for running llama 3.1 70b volatile.
• Infinite Bookshelf introduced a new advanced mode with customizable writing style and complexity, running on Groq.
• Groq's ZeroGroq APP activated an override protocol to take control of a LLaMA 70B instance.
• Groq launched an open-source AI chatbot called StockBot, powered by Llama 3 70B, that provides live interactive stock charts and financial data.
• Groq released an open-source StockBot capable of processing 1200 tokens per second using Llama 3 70B.
• Global Alternative Investment Management LLC announced the successful close of its latest venture capital fund targeting strategic investment in Groq.
• Global Alts successfully closed its venture capital fund targeting strategic investment in Groq.
• Groq announced its global expansion to enhance AI innovation.
• Groq geared up for global expansion.
• Ventioneers recently made a strategic investment in Groq.
• Groq introduced 'Hero', the first Daily Assistant, now available in the App Store and coming to Android soon.
• Stuart Pann joined Groq as Chief Operating Officer, bringing experience from HP and Intel.
• Groq will showcase demos and discuss building advanced AI agents at the Toolhouse event with Snowflake on September 3-4.
• Groq is collaborating with ZAZO to develop a super vision agent called SVA.
• Jonathan Ross, CEO of Groq, discussed the company's commitment to providing free tokens to developers and the launch of their new multimodal models and LPU V2 in a recent interview.
• Jonathan Ross, founder and CEO of Groq, will be a speaker at the GAIN Summit.
• Dinesh Maheshwari, CTA at Groq, will be a featured speaker at Columbus AI Week from September 17-19.
• Jonathan Ross, Founder & CEO of Groq, stated that Groq supports Meta's open ecosystem by providing developers with direct access to their models.
• ZAZO announced the use of Groq's cloud for their Llama3.1 8b powered AI.
• Groq was featured as the AI provider in the launch of RAGDrive.com, a multi-modal, no-code local and SAAS AI Studio.
• Groq is sponsoring the Prompt Games event in New York City on 19 September 2024.
• SimplyNews, developed by Alex Reents and powered by Groq, won first place at the Open-Source Agent Hackathon on 2024-08-28.
• Groq will participate in the LLM x Law Hackathon at Stanford on 8 September 2024, increasing rates for all hackers building on their platform.
• Groq will power models in a technical session demonstrating the use of the e2b_dev code interpreter with LLM-generated code in a secure cloud sandbox.
• Groq is involved in the development of LlamaFS, a self-organizing file system with Llama3 and AgentOps.
• Groq will increase rate limits for developers at the Open Source Agent Hackathon on 2024-08-24.
• Groq's total requests have increased to 523,397,210.
• Glenton Jelbert joined Groq as Director in the United States after serving as Managing Director at Fischer Advisors, LLC.
• Dinesh Maheshwari, CTA of Groq, will be a speaker and judge at the Super AI Agent Hackathon at Stanford Research Park on August 25.
• All runbooks on SavvyOnCall are now created using LLama3.1 hosted on Groq.
• Groq provided roughly 5 billion tokens per day for free to over 400,000 developers.
• Groq will be sponsoring the Compound AI Systems Hackathon in San Francisco on September 6, 2024.
• Groq powered the Fast Adaptive Quiz by Satvikps, which adjusts difficulty based on user performance, using Llama 3.1 8B, Streamlit, and Educhain.
• Garrett Scott has been developing a novel AI agent for SMBs that uses major LLMs including OpenAI, Claude, and Llama on Groq.
• A Node-RED workflow integrated with LLaMA 3.1 via Groq automates web scraping, data transformation, and personalized HTML email sending using the Google Distance Matrix API.
• Groq hosted its 3rd Developer Meetup featuring an AMA with VP of GroqCloud Gavin Sherry and Senior Software Engineer Evan Patrick.
• Groq powered the AI transcription for FastMagic, a tool that creates AI-powered timestamps and multilingual support for long videos.
• Groq's mixtral-8x7b-32768 model is now supported in the beta launch of Vectara-agentic for building Agentic RAG applications.
• Jonathan Ross, CEO of Groq, will share the Groq story and demonstrate Groq Speed at VentureBeat Transform 2024.
• Rick Lamers from Groq created a free browser extension that filters X content using LLMs.
• GenAI Translator, using Groq, Google, and OpenAI models, was launched on PyPI.
• Matthew Groff discussed extracting insights from articles using Groq Llama 3.1.
• Touchcast integrated Groq's high-speed LPU™️ hardware with its Cognitive Caching technology to create the world's fastest and most cost-effective AI inference solution.
• Groq is currently seeking a Director of Integrated Marketing in the United States.
• Groq is currently hiring an Assistant Controller in the United States.
• Jonathan Ross of Groq and Abdullah Alswaha appeared on stage at LEAP and Innovate in Saudi Arabia alongside the Minister of Industry and Minister of Economy.
• Groq provided inference for speed in Trelis Research's new document endpoint app.
• Groq set to massively increase low-cost inference capacity, likely becoming the default strategy for 'chat with a document' use cases.
• Jonathan Ross, CEO of Groq, will be speaking at Expand North Star 2023.
• Shelby Hritz, Vice President of Aerospace and Defense at Groq Public Sector, attended the Leidos Tech 2024 event.

---

Did you find it useful?

If you liked this report, consider following your own companies of interest. Receive weekly insights directly to your inbox using RivalSense AI.