Head to head: Sarvam M vs DeepSeek-V4-Pro
DeepSeek-V4-Pro defeated Sarvam M 113.0 to 44.4 in a 12-task benchmark, achieving a 12-0 sweep with 100% confidence. The test, judged twice by gpt-5.4 to cancel position bias, found DeepSeek-V4-Pro re…
DeepSeek-V4-Pro defeated Sarvam M 113.0 to 44.4 in a 12-task benchmark, achieving a 12-0 sweep with 100% confidence. The test, judged twice by gpt-5.4 to cancel position bias, found DeepSeek-V4-Pro re…
In a head-to-head test of eight GSM8K math word problems, gpt-oss-120b and Kimi K3 tied on seven tasks, both returning the correct numeric answer every time. The only separation came on the eighth tas…
Jason Zhao announced on July 22, 2026, via his personal X account the launch of WhatIfVerse, an AI-driven simulation platform where virtual citizens work, earn wages, and interact in a continuously ev…
RuntimeWire launched a public MCP server on July 22nd, 2026, that gives AI agents unrestricted real-time access to its AI-economy news wire without requiring an API key or signup. The service streams …
Ema, an enterprise AI startup led by Surojit Chatterjee, is in talks to raise about $80 million in a round led by Creaegis at a valuation of roughly $800 million, according to a Moneycontrol report ci…
Ex-Rapyd leaders Eric Rosenthal, Lior Levit, Sagi Ittah and Juan Jose Huezo raised $8 million in seed funding co-led by Motive Partners and Oak HC/FT to launch Cordant, a payment observability layer t…
YC-backed Prescience launched on July 21st, claiming its AI-managed health benefits platform can cut employer premiums by roughly 20% while offering $0-deductible plans. Founded by brothers Aditya Jai…
GMI Cloud, led by Alex Yeh, reported that signed customer commitments pushed contracted annual recurring revenue above $500 million as of July 21st, while live ARR grew 2.4x during the first half of 2…
Adaption launched Teams on July 21st, adding shared compute, centralized billing and member-level controls to its model-building platform. The release turns collaboration and spending controls into pr…
Base44, a Wix-owned startup founded by Maor Shlomo, launched a seven-day developer competition on July 21st offering $10,000 to one winner for an application built through an external IDE or AI coding…
RuntimeWire launched the beta of RuntimeWire New Tab, a free Chrome extension built on Manifest V3 that replaces Chrome's blank new-tab page with an AI-news dashboard featuring a slide-out feed, daily…
Nanbeige released Nanbeige4.2-3B, a three-billion-parameter language model using a Looped Transformer architecture that the company says increases model capacity without adding parameters, delivering …
SpaceXAI opened applications Tuesday for an August 8th Grokathon in San Francisco, giving selected developers access to its Grok 4.5 model and X APIs less than a month after releasing the model. The 1…
ParcoData, a nuclear-powered AI data center developer, has sold $544,910 of equity in its first disclosed financing toward a $2.31 million offering, according to a July 21st SEC filing. The Miami Beac…
Antim Labs opened the public beta of Gizmo on July 21st, an AI agent that converts text and reference images into editable 3D environments for robot training. The tool generates structured assets with…
Cascade, an AI platform co-founded by Hannia Zia and Joana Ferreira, has raised $3.5 million from Andreessen Horowitz Speedrun, Ada Ventures, and others to help architecture, engineering, and construc…
TwelveLabs announced on July 21st, 2026, that its first video-intelligence AI agent, Jockey, is now available in a research-preview program, enabling users to search entire video libraries via Anthrop…
Libretto, a two-person San Francisco startup founded by Michael Kronovet and Tanishq Kancharla, released Browser Tools SDK, a package of six Playwright-based tools designed to reduce token usage and r…
Intermezzo, a Menlo Park payroll infrastructure developer, reported $8.3 million in equity sales as part of a $10.3 million private offering in a Form D filed with the SEC on July 21st. The company, f…
BlueFolder launched two optional AI features on July 21st that summarize customer histories and technician notes inside its field-service platform, giving dispatchers and managers a faster way to read…