Foundation Models - News Articles

Google Unveils Gemini 4 Argon, Rolls Frontier Model Out to Cyber Defenders

Oct 01, 2026

The Deep View

Google unveils Gemini 4 Argon, a frontier model now rolling out to cyber defenders through its restricted Fairwind program, with a 1 million-token output limit and introductory pricing of $2 per million input tokens and $10 per million output tokens.

Foundation Models Security

Google Unveils Gemini 4 Argon, Claims Lead Over Rivals in Coding, Knowledge Work and Cybersecurity

Oct 01, 2026

The Deep View

Google unveils Gemini 4 Argon, a flagship AI model it claims leads rivals in coding, knowledge work and cybersecurity, with restricted Fairwind access for cyber defenders and pricing starting at $2 per million input tokens.

Foundation Models Coding & Developer Tools Security AI Safety

CrowdStrike Says AI Shifts Cybersecurity Balance Toward Defenders as It Helps Disrupt Sality Botnet

Oct 01, 2026

The Deep View

CrowdStrike says AI is beginning to shift cybersecurity’s advantage toward defenders as it and partners disrupt the two-decade-old Sality botnet, using Nvidia Nemotron-based red- and blue-team models trained on CrowdStrike security data.

Security Foundation Models

Google Unveils Gemini 4 Argon, Giving Trusted Cyber Defenders a Model It Says Can Autonomously Patch Critical Flaws

Oct 01, 2026

Google

Google unveils Gemini 4 Argon, a cyber-defense model it says can autonomously find, validate and patch critical vulnerabilities, initially restricting access to trusted defenders in its Fairwind Program after Wiz uses it to uncover a healthcare-software flaw earlier frontier models missed.

Security Foundation Models Coding & Developer Tools

Google Unveils Gemini 4 Argon for Select Cybersecurity Teams, Claiming Benchmark Lead Over GPT-6 Astra and Claude Opus 5.5

Oct 01, 2026

The Rundown AI

Google unveils Gemini 4 Argon for vetted cybersecurity teams, claiming it leads GPT-6 Astra and Claude Opus 5.5 on 13 of 19 internal benchmarks and reaches 77.9% on DeepSWE coding.

Foundation Models Security Coding & Developer Tools

Fig launches RIDGE, finding frontier AI models have jagged task performance with no clear overall leader

Sep 30, 2026

Fig.inc

Fig launches RIDGE, finding no clear frontier-AI champion as Astra, Opus 5.5 and rivals show jagged task-level results; Astra leads VisualWebArena at 139 of 177 tasks, yet every lower-scoring model solves some tasks it misses.

Foundation Models Research Agents

PageIndex SDK Adds Local Indexing, Retrieval and Chat With Users’ Own LLM Keys

Sep 30, 2026

GitHub

PageIndex SDK now lets developers index, retrieve from, and chat with text-based PDFs locally using their own LLM keys while retaining access to PageIndex Cloud through the same client, with PageIndex Flash becoming its default local tree-indexing method.

Coding & Developer Tools Data & Analytics Foundation Models

OpenAI unveils Dots always-on AI agents and GPT-6.1 Sol at DevDay

Sep 30, 2026

The Deep View

OpenAI unveils its Dots strategy of always-on agents running on separate cloud computers and launches GPT-6.1 Sol, which it says approaches Astra-level intelligence for $2 per million input tokens, while introducing a $500 monthly Ultrafast tier and halving some $200 Pro allowances.

Agents Foundation Models

OpenAI launches always-on ChatGPT Dots agents and GPT-6.1 Sol

Sep 30, 2026

The Deep View

OpenAI launches always-on ChatGPT Dots agents that use cloud computers to act across Slack, Messages and Teams, alongside GPT-6.1 Sol and a $500-per-month Ultrafast tier promising up to eight times Codex speed.

Agents Consumer Products Foundation Models Coding & Developer Tools

OpenAI Launches Always-On ChatGPT Dots Agents That Work 24/7 Across 4,000 Apps

Sep 30, 2026

The Rundown AI

OpenAI launches Dots, always-on ChatGPT agents powered by GPT-6 Astra that run from cloud computers, connect to more than 4,000 apps, and can answer in Slack, Teams, and ChatGPT 24/7.

Agents Enterprise Cloud Computing Foundation Models