Groq has been recognized as a Cool Vendor in the 2025 Gartner AI Infrastructure report, highlighting its LPU chip's deterministic, low-latency inference that scales linearly. Over 2.5 million developers use Groq for up to 5x faster and cheaper performance than GPUs.

The article discusses U.S. leadership in AI compute, especially inference, and proposes an export policy that balances market flexibility with consortium coordination to maintain strategic advantage.

GroqCloud is expanding its AI inference infrastructure globally to meet the growing demand from real-time applications moving from experimentation to production. A new UK data center, in partnership with Equinix, brings deterministic, high-performance inference closer to European developers and enterprises. GroqCloud now has over 3.5 million developers and sustained increases in production traffic.

Groq’s LPU is purpose-built for inference, achieving ultra-low latency without sacrificing accuracy through TruePoint numerics, SRAM-based memory, static scheduling, and tensor parallelism. Kimi K2 runs at 40x performance on Groq, demonstrating the architecture’s efficiency.

Anthropic announced a $200 million partnership with the Gates Foundation to provide grants, Claude usage credits, and technical support over four years, focusing on global health, life sciences, education, and economic mobility. The collaboration aims to leverage Claude AI to accelerate vaccine development, improve healthcare decisions, enhance educational outcomes, and boost agricultural productivity.

Anthropic launches Claude for Small Business, integrating AI into tools like QuickBooks, PayPal, and HubSpot with 15 pre-built workflows to automate finance, marketing, and operations. Includes free AI fluency course, roadshow, and partnerships with CDFIs.

An exploration of how AI-assisted programming leads to burnout and cognitive debt, with insights from industry experts on the need for sustainable practices and understanding code structure.

ElevenLabs has surpassed $500M ARR in early 2026, up from $350M at end of 2025, driven by enterprise adoption of voice agents. New investors include BlackRock, NVIDIA, Jamie Foxx, and Eva Longoria, among others. The company plans to expand its platform and international presence.

Anthropic announced a partnership with SpaceX to gain over 300 megawatts (220,000 NVIDIA GPUs) of compute capacity from the Colossus 1 data center, effective immediately, along with increased usage limits for Claude Code and the Claude API. The company also highlighted other major compute agreements with Amazon, Google/Broadcom, Microsoft/NVIDIA, and Fluidstack, and outlined international expansion plans focusing on compliance and data residency.

During the 2025 Polish Presidency of the Council of the European Union, the General Secretariat partnered with ElevenLabs to dub high-level diplomatic press conferences into English, French, and Polish, making them accessible to millions of EU citizens. The collaboration emphasized authenticity, human-verified translations, and strict privacy protocols.

Deutsche Telekom partners with ElevenLabs to bring AI voice agents to customer service, offering 24/7 human-like interactions with no waiting times.

Klarna has launched a voice AI agent built with ElevenAgents, live as the first line of phone support for customers in the United States. The assistant resolves queries up to 10x faster, reducing wait times and freeing up human agents. Plans are underway to expand globally to 114 million customers.

ElevenLabs announced a $500 million Series D funding round, valuing the company at $11 billion, more than triple its valuation from a year ago. With over $330 million in ARR in 2025, the company will use the funds to expand its enterprise voice agent platform ElevenAgents, invest in research, and support global expansion.

ElevenLabs opens a Madrid office and expands operations in Spain, building local sales and engineering teams to support enterprise customers deploying AI agents. Examples include eDreams ODIGEO, MediaMarkt, and Santa Lucia Seguros. Santander has invested. The company highlights Spain's linguistic diversity as an opportunity for voice AI.

ElevenLabs launches ElevenMusic, an AI-powered platform for music discovery and creation. Users can discover music, remix tracks, or create from scratch, and earn from listener engagement. The platform launches with over 4,000 independent artists and Eleven Album Vol. 2.

ElevenLabs now offers on-premise and on-device deployment, expanding options beyond cloud and VPC for various enterprise environments.

Kimi K2 is an open agentic intelligence platform featuring tools for Excel formulas, document conversion, AI agent deployment, code assistance, and research advancements like Kimi K2.6 and Agent Swarm.

Kimi K2 is an open-source thinking model with tools for Excel, docs, code, and web browsing, supporting agent swarms and deep research.

Kimi open-sources the Vendor Verifier (KVV) to help users verify the accuracy of inference implementations of open-source models. It includes six critical benchmarks for detecting common deployment issues and encourages infrastructure providers to fix root causes.

Kimi K2.5 is an open-source multimodal model that delivers state-of-the-art performance in coding and vision tasks. It features a self-directed agent swarm capable of orchestrating up to 100 sub-agents for parallel execution, reducing task completion time by up to 4.5x. The model also excels in office productivity, handling complex documents, spreadsheets, and presentations. Available on multiple platforms, Kimi K2.5 represents a significant step toward AGI for the open-source community.

WorldVQA is a new benchmark to evaluate factual correctness of MLLMs on visual world knowledge. It includes 3,500 high-quality image-question pairs across 9 categories, with a focus on head vs tail distribution. Frontier models achieve below 50% accuracy, revealing overconfidence and gaps in visual knowledge.

Kimi launches Agent Swarm, a multi-agent architecture enabling up to 100 parallel sub-agents for horizontal scaling. The system self-organizes into roles like CEO, researcher, and analyst, autonomously decomposing tasks, assigning agents, and synthesizing results. It is up to 4.5x faster than sequential execution and excels in broad research, batch processing, and multi-perspective analysis. Now available in preview for top-tier subscribers.

Kimi K2.6 is a new open-source model with state-of-the-art coding, long-horizon execution, and agent swarm capabilities. This blog details its features, benchmarks, and community feedback.

DeepSeek has released R1-0528, featuring improved benchmark performance, enhanced front-end capabilities, reduced hallucinations, and support for JSON output and function calling. The model is available on the chat platform with no API changes, and weights are open-sourced.

DeepSeek has released V3-0324 with major improvements in reasoning, front-end development, and tool-use capabilities. The model is now under the MIT License, and API usage remains unchanged.

DeepSeek released the R1 model on January 20, 2025, with performance on par with OpenAI-o1, fully open-sourced under MIT License, along with distilled small models and API access.

DeepSeek App launched on 2025/01/15, powered by DeepSeek-V3, free with no ads or in-app purchases. Available on App Store, Google Play, and major Android markets. Features include easy login, cross-platform sync, web search, Deep-Think mode, and file upload.

DeepSeek-V3 is the biggest leap forward yet, featuring 671B MoE parameters (37B activated), 60 tokens/second (3x faster than V2), and training on 14.8T high-quality tokens. API pricing is same as V2 until Feb 8, then $0.27/M input (cache miss), $0.07/M (cache hit), $1.10/M output. Fully open-source models and papers.

DeepSeek releases V2.5-1210 as the final version of the V2.5 series, introducing internet search, improved benchmarks in math, coding, writing, and roleplay, and open-sourcing the model on Hugging Face. The team thanks users and hints at next-gen foundation models.

DeepSeek launches R1-Lite-Preview, achieving o1-preview-level performance on AIME & MATH benchmarks, with real-time transparent reasoning and upcoming open-source release.