AI News: June 4, 2026

1. AI Music Startup Suno Raises $400 Million at a $5.4 Billion Valuation

Suno. AI music generator Suno raised a $400 million Series D at a $5.4 billion valuation, roughly doubling its value from seven months earlier, in a round led by Bond Capital with IVP, Forerunner, Union Square Ventures, Lightspeed, and Menlo Ventures. The company reports more than two million subscribers and projects $300 million in annual revenue, even as it faces copyright litigation from Universal Music Group and Sony Music over training data. The raise shows investors still backing generative-audio companies despite unresolved legal exposure over music rights. Source

2. DeepSeek Nears $7.4 Billion Raise at a $45 Billion Valuation

DeepSeek. Chinese AI lab DeepSeek is closing in on a roughly $7.4 billion funding round at an approximately $45 billion valuation, one of the largest startup financings in China to date, with backing from Tencent, the government-backed National Artificial Intelligence Industry Investment Fund, and founder Liang Wenfeng. The company says it will prioritize fundamental AI research and open-source models over near-term commercialization. The scale of the round underscores the continued state-aligned capital flowing into Chinese open-weight model development. Source

3. Coralogix Raises $200 Million to Monitor AI Agents

Coralogix. Observability startup Coralogix raised a $200 million Series F at a $1.6 billion post-money valuation, led by Advent and the Canada Pension Plan Investment Board, bringing total funding to $550 million. The company, which analyzes logs, metrics, and traces for over 5,000 customers including IBM and JFrog, says more than half of its large enterprise clients now interact with its platform through an AI agent or their own models rather than dashboards. The round reflects investor appetite for the monitoring layer needed to operate autonomous agents in production. Source

4. Lovable Signs Multiyear Google Cloud Deal to Expand Capacity Fivefold

Lovable. European AI coding platform Lovable signed a multiyear agreement with Google Cloud that increases its footprint and AI usage roughly fivefold, expands access to both Anthropic’s Claude and Google’s Gemini models, and lists its agents on Google Cloud’s enterprise marketplace. Lovable, which reports over $400 million in annualized revenue, gains enterprise distribution while Google drives consumption of Anthropic services it has invested in. The deal illustrates how hyperscalers are locking in fast-growing AI application companies through compute and distribution commitments. Source

5. AethexAI Raises $3 Million for Voice AI in Africa and the Middle East

AethexAI. AethexAI, founded by ex-Goldman Sachs executive Mariama Diallo and former Meta engineer Ayooluwa Odemuyiwa, raised $3 million in pre-seed funding led by 4DX Ventures to build voice AI customer-support systems for African and Middle Eastern markets. Backers include Enza Capital, Dorm Room Fund, Mojo Ventures, and individual Anthropic researchers. The round highlights growing investment in localized voice agents for languages and regions underserved by mainstream providers. Source

6. Nous Research Releases Hermes Desktop, an Open-Source Cross-Platform AI Agent

Nous Research. Nous Research released Hermes Desktop, an MIT-licensed open-source AI agent app in public preview for Windows, macOS, and Linux that integrates with Telegram, Discord, Slack, WhatsApp, Signal, email, and the terminal. It offers persistent memory, natural-language task planning, sub-agent delegation, and five sandboxed execution backends (local, Docker, SSH, Singularity, and Modal), with access to more than 300 models through the Nous Portal. The release gives developers a freely modifiable, model-agnostic alternative to closed desktop agents. Source

7. Ideogram 4.0 Launches as a Leading Open-Weight Image Model

Ideogram. Ideogram released version 4.0 as an open-weight text-to-image model with native 2K resolution, improved text rendering for logos and posters, and precise layout control via bounding boxes, with weights and code on GitHub under a license that requires paid commercial use. The company says it ranks first among open-weight image models on industry benchmarks, competing closely with proprietary systems from OpenAI and Google. The release narrows the gap between open and closed image generation for developers running models on their own hardware. Source

8. Alibaba Launches Qwen3.7-Plus With Vision and Agentic Capabilities

Alibaba. Alibaba’s Qwen team launched Qwen3.7-Plus, a multimodal model that understands images and video alongside text and adds agentic features including deep reasoning, self-programming, tool invocation, verification, and autonomous iteration. The model is available to external developers through Alibaba Cloud’s Bailian platform, marketed internationally as Model Studio. It extends the Qwen family’s push toward multi-step agentic execution rather than single-turn responses. Source

9. Clarius Gets FDA Clearance for an Ejection Fraction AI Tool

Clarius Mobile Health. Clarius received FDA 510(k) clearance for its Ejection Fraction AI, a semiautomated tool that calculates and displays left ventricular ejection fraction in real time from standard cardiac windows on Clarius wireless ultrasound scanners. The system was trained on thousands of clinician-annotated cardiac ultrasound images. The clearance adds to a steady stream of AI-enabled point-of-care imaging tools moving cardiac measurement closer to the bedside. Source

10. Tempus AI Receives FDA Clearance for Tumor-Only xT CDx Use

Tempus AI. Tempus AI received FDA clearance for tumor-only use of its xT CDx genomic profiling platform, alongside new AI-driven diagnostics, broadening the assay’s clinical applicability in oncology. The clearance lets clinicians run the companion diagnostic without a matched normal sample, simplifying testing workflows. It marks continued regulatory expansion for AI-supported precision-oncology diagnostics. Source