Baseten model performance engineer Michael Feil shipped Baseten Tokenizer, a Rust-backed package that cuts CPU time for preparing long prompts for Moonshot AI's Kimi K3 model, achieving up to 18x fast…
Baseten has launched day-0 API support for Kimi K3, a new open frontier model from Moonshot AI with 2.8 trillion parameters, making it the largest open model to date. The API runs on NVIDIA GB300 NVL7…
Moonshot AI's Kimi K3 and Kimi K3 Fast models are now available on Vercel's AI Gateway from US-based providers including Baseten and Fireworks, with Zero Data Retention (ZDR) support. The models run o…
Baseten leads the public GLM-5.2 inference speed table with 602.8 output tokens per second, roughly 31% faster than Blackbox AI at 460.4 tokens per second, according to Artificial Analysis median meas…
Baseten has built the fastest API for GLM-5.2, achieving peak speeds of 280 tokens per second and average speeds around 100 tokens per second, more than double the performance of the launch-day API as…
Sarah Guo, who became Greylock's youngest general partner at 28 and left four years later to found Conviction, staked her firm entirely on AI before ChatGPT launched, with pre-ChatGPT seed checks into…
Gas prices have risen above $4 per gallon, with potential for further increases due to developments in the war with Iran, according to Matt Smith, Director of Commodity Research at Kpler. Meanwhile, C…
Inference startups Fireworks, Baseten, and Together AI collectively raised $3.8 billion in four weeks, signaling that inference has become a control point in the AI industry, according to Forbes contr…
Nvidia on July 1 announced software optimizations that boost token throughput by up to 5x for DeepSeek V4 on Blackwell systems, with throughput improvements reaching 20x compared to baseline configura…
NVIDIA announced a new financing model combining revenue sharing with credit support to help AI cloud providers build GPU infrastructure for startups and enterprises. The initiative, detailed in a Jul…
A developer's benchmark of frontier AI models on a 3D modeling task found Anthropic's Opus 4.7 outperformed Gemini 3.1, GPT 5.5, and Kimi K2.6 in weighted score, speed, and cost, contradicting public …
Nvidia and LangChain released the NemoClaw for LangChain Deep Agents blueprint on July 8, combining Nvidia's Nemotron 3 Ultra model with LangChain's Deep Agents framework to achieve benchmark-leading …
NVIDIA and LangChain released a tutorial on creating a LangChain Deep Agents harness profile for NVIDIA Nemotron 3 Ultra to match proprietary frontier model intelligence. The profile optimizes agent p…
NVIDIA is launching a revenue-sharing and credit-support model with AI cloud partners to accelerate large-scale AI factory deployments. The model allows partners to build and operate DSX AI factories …
NVIDIA launched a new business model enabling AI clouds to procure its infrastructure through revenue-sharing and credit-support, accelerating access to large-scale accelerated computing for startups,…
A developer built Recowork, a desktop agent application inspired by Claude Desktop, using Tauri, OrbStack, and Apple container machines with GLM-5.2. The app plans, calls tools, and chains steps to co…
South Park Commons portfolio companies raised over $1.5 billion in June 2026, including Baseten's $1.5B Series F at a $13B valuation for AI inference and Orb's $335M acquisition by Adyen. AXION launch…
NVIDIA's full-stack inference software, codesigned with its hardware, has reduced token costs by up to 5x on the DeepSeek V4 model in one month on the Blackwell platform. Companies like Baseten, Cogni…
A new benchmark of 14 API providers for the GLM-5.2 (max) model reveals Fireworks leads in output speed (261.5 t/s) and low latency (9.81s), while GMI (FP8) offers the lowest blended price at $0.72 pe…
Baseten has built the world's fastest API for GLM-5.2, achieving over 280 tokens per second as measured by Artificial Analysis. The performance is driven by optimizations including an updated inferenc…