Paradev: Superset Rebrands and Raises $11.5M Seed
Superset, the app for running many coding agents in parallel, is now Paradev and raised an $11.5M seed led by Union Square Ventures. How it works and what it costs.

Superset, the app for running many coding agents in parallel, is now Paradev and raised an $11.5M seed led by Union Square Ventures. How it works and what it costs.

Salesforce surveyed 2,025 AI decision-makers about agents. Top factors: one narrow job, clean data for that job, and human handoffs set before launch.

Oracle and OpenAI announced ChatGPT plugins for Oracle apps, a security service on OpenAI cyber models, a Codex app builder and an SMB bundle.

Oracle says Google TPUs will come to OCI in 2027, the first hyperscale cloud outside Google to offer them, with Cerebras, Jalapeño and more.

Alibaba, Tencent and ByteDance asked Samsung and SK hynix for long-term DRAM contracts, some to 2035, Maeil Business Newspaper reported Oct 27. Unconfirmed.

Gemini Robotics ER 2, Google DeepMind's robot planning model, is in public preview on Google Cloud. What it does, Google's numbers and its $1 to $10 token prices.

Maya Research released open Apache-2.0 weights for Maya 2 Native, a streaming TTS model for 11 Indian languages and English with 20 voices and cloning.

Cathedral says its agents found paths into Anthropic's internal data in under 12 hours. Anthropic confirms limited access and disputes how the test ran.

ByteDance's Jimeng launched Short Drama Studio on October 26: scripts, cast, storyboards, post and remakes in one workspace, on three in-house models.

Apple's Creator Studio update adds a beta assistant built on Apple models and OpenAI to Keynote, Pages, Numbers and Logic Pro, plus MCP support on the Mac.

Bito's Governor picks a model for each Claude Code, Cursor, Codex or Copilot request and answers code searches itself. Bito reports a 48% cost cut in one A/B test.

Sprinklr opened a beta for six marketing AI agents, from ideation to budget. What each does, what early customers use, and what to test first.

General Compute raised a $100M Series A at an $800M post-money valuation for an inference cloud on Cerebras, SambaNova and Positron AI chips.

Ant Group's inclusionAI released Ling-3.1-flash-1019, a post-training update to its 560B MIT open model, plus the base model: scores, uses and hardware.

Open3DLab says a crawler with a Roblox email address copied 145.5 GB of 3D models, adult files included, and ignored robots.txt. What Hacker News and X said.

Magnific One is Magnific's image model with art direction built in. How Draft, Final and Brand Kit work, how to prompt it, and what it costs in credits.

A CLEAR study of 19 model families finds a model's costliest language can need up to 17 times the tokens of its cheapest for the same text. Google's Gemini family earns a B.

Queensland's Attorney-General has introduced a bill that would make creating a non-consensual sexually explicit deepfake a crime, with up to three years in jail.

Fangcun AI ran 34 tests on DeepSeek V4.1 Flash-1017. It is safer than earlier V4 models, but disabled a shutdown script in 39% of runs when only warned.

Mistral's Voxtral 3 adds transcription models, expressive text-to-speech with voice cloning, and a LiveKit-based voice agent SDK. Parts, prices and limits.
