Recent posts

Daniel Kowalski Industry
3 hours ago

OCRia-1 Sets New Document Parsing Benchmarks

LlamaIndex has released OCRia-1, a 4B vision-language model for document parsing.

Daniel Kowalski Industry
4 hours ago

U.S. Banks Restrict AI Agents Over Account Security Risks

Major U.S. banks are restricting third-party AI agents after detecting automated access to customer accounts.

Photo by Magda Ehlers on Pexels
Julie Davis AI Models
1 day ago

Anthropic Previews Claude Duet Shopping Flow With Stripe Link

Anthropic has shown an early look at Claude Duet, an upcoming proactive agent, with an agentic shopping flow using Stripe Link.

Photo by Tara Winstead on Pexels
Sergey Bauer Industry
1 day ago

ElevenLabs Releases Eleven Realtime v1 Max and Mini

ElevenLabs launched Eleven Realtime v1 Max and Eleven Realtime v1 Mini, two low-latency speech-to-speech models available through ElevenAgents and the Realtime API.

Photo by Jessica Lewis 🦋 thepaintedsquare on Pexels
Peter Hale Industry
2 days ago

Anthropic’s $600B Compute Bet Highlights AI Infrastructure Demand

Anthropic is reportedly committing more than $600 billion to future cloud and AI hardware capacity after generating about $16.3 billion in revenue in 1H26 on roughly $9 billion of compute spending.

Photo by Nicolas Foster on Pexels
Alex Novak Industry
2 days ago

NTT Plans Optical Chiplet Production Line for AI Data Centers

NTT is preparing a production line in Japan for optical chiplets that use light instead of electrical signals to move data between CPUs, GPUs and other processors.

Photo by Andrey Matveev on Pexels
Sergey Bauer Industry
3 days ago

SKY-001: Engineering World Model Targets CAD and BIM

Skyfall AI has introduced SKY-001, a specialized model for CAD, BIM and physical-object design.

Photo by Kindel Media on Pexels
Julie Davis Industry
3 days ago

Google and Cognizant Expand Gemini AI Partnership With 18,000-Engineer Group

Google Cloud and Cognizant are expanding their AI partnership with a new Gemini Frontier Capability Group involving more than 18,000 Cognizant engineers and specialists.

Photo by BM Amaro on Pexels
Julie Davis AI Models
1 week ago

Brave Leo Gets End-to-End Encrypted AI Chat With GLM-5.3 Flash

Brave has added GLM-5.3 Flash to Leo with end-to-end encrypted inference based on a Trusted Execution Environment (TEE).

Photo by Tim Witzdam on Pexels
Peter Hale AI Models
1 week ago

NVIDIA Nemotron 3.5 Super VL: Specs and Differences vs Nemotron 3 Super

NVIDIA Nemotron 3.5 Super VL is a 120B-parameter open multimodal model with 12B active parameters, built for long-running AI agents. It supports reasoning, tool use, and native understanding of text, images, documents, and video, while keeping the 1M-token context window of Nemotron 3 Super.

Photo by UMA media on Pexels
Daniel Kowalski Startups
1 week ago

AIPOCH Open-Science Introduces Portable .science Research Packages

AIPOCH Open-Science is an open-source research environment built around reproducibility, local execution and portable research sessions.

Photo by Armin Rimoldi on Pexels
Sergey Bauer Startups
1 week ago

OpenAI UK Revenue Rose 270% in 2025, Profit Reached £20.3 Million

OpenAI UK Ltd reported £292.96 million in turnover in 2025, up from £79.21 million in 2024. Revenue increased by approximately 270% year over year.

Photo by Kindel Media on Pexels
Peter Hale AI Models
1 week ago

Anthropic Warns AI Worms Could Soon Spread Across Smart Devices Without Human Control

Anthropic researchers say small AI models are approaching the capability needed to copy themselves between internet-connected devices.

Photo by panumas nikhomkhai on Pexels
Julie Davis AI Models
1 week ago

Workday Connects Sana AI Agents With Claude

Workday and Anthropic announced an integration between Sana AI agents and Claude for HR and finance workflows.

Photo by Sanket Mishra on Pexels
Daniel Kowalski Industry
1 week ago

Google Fitbit Edge launched: specifications and comparison

Google has launched the Fitbit Edge for $179.99. The new fitness tracker features a 1.34-inch AMOLED display, built-in GPS, up to 7 days of battery life, health tracking, and Gemini-powered personalised coaching.

Photo by Vitaly Gariev on Pexels
Julie Davis AI Models
1 week ago

Gemini 4 Argon Could Roll Out More Widely Next Week

Google may expand access to Gemini 4 Argon as soon as next week.

Photo by Tara Winstead on Pexels
Julie Davis AI Models
2 weeks ago

Google brings Gemini and Auto Browse to Chrome on Android in India

Google has rolled out Gemini in Chrome for Android users in India. The desktop version launched in the country earlier in 2026.

Photo by Pixabay on Pexels
Sergey Bauer AI Models
2 weeks ago

EmbeddingGemma 2: technical overview and comparison with EmbeddingGemma

EmbeddingGemma 2 is Google’s lightweight multimodal embedding model for text, code, images, video, and audio. It supports up to 740M parameters, an 8K context window, 128–768 dimensional embeddings, and local on-device deployment under the Apache 2.0 license.

Photo by Ron Lach on Pexels
Julie Davis AI Models
2 weeks ago

OpenAI Adds Invisible Watermarks to ChatGPT and Codex Text

OpenAI is rolling out invisible text watermarks for ChatGPT and Codex in the EU. API customers worldwide can opt in for supported models.

Photo by Kawê Rodrigues on Pexels
Alex Novak Industry
2 weeks ago

eFlesh: NYU Open-Sources a 3D-Printed Magnetic Tactile Sensor for Robotics

Researchers at New York University introduced eFlesh, an open-source tactile sensing platform for custom robotic hardware.

Photo by FOX ^.ᆽ.^= ∫ on Pexels
Posts feed
OCRia-1 Sets New Document Parsing Benchmarks
3 hours ago
U.S. Banks Restrict AI Agents Over Account Security Risks
4 hours ago
Anthropic Previews Claude Duet Shopping Flow With Stripe Link
1 day ago
ElevenLabs Releases Eleven Realtime v1 Max and Mini
1 day ago
Anthropic’s $600B Compute Bet Highlights AI Infrastructure Demand
2 days ago
NTT Plans Optical Chiplet Production Line for AI Data Centers
2 days ago
SKY-001: Engineering World Model Targets CAD and BIM
3 days ago
Google and Cognizant Expand Gemini AI Partnership With 18,000-Engineer Group
3 days ago
Brave Leo Gets End-to-End Encrypted AI Chat With GLM-5.3 Flash
1 week ago
NVIDIA Nemotron 3.5 Super VL: Specs and Differences vs Nemotron 3 Super
1 week ago

Seatofish.com publishes content that serves only as information and education about mobile technology, apps, and artificial intelligence. If you act on any material that appears on this website, you accept full responsibility for the outcome. We urge every reader to perform independent research when necessary.

Seatofish.com operates as a self contained media outlet - it reports on mobile technology, iOS and Android platforms, apps, gadgets, and artificial intelligence.
info@seatofish.com

© 2012-2026 seatofish.com

We use cookies to improve your experience on our site and to show you relevant advertising. To find our more, read our privacy policy and cookie policy