Pillar · Models

Frontier AI Models

GPT-5, Claude Opus, Gemini Ultra, Llama 4, Grok, Mistral, DeepSeek, Qwen. Releases, benchmarks, pricing, and what each one is actually good at.

Newest first

All Frontier AI Models coverage

Last updated
Editorial render of the Claude Opus 4.8 release with a Dynamic Workflows subagent fanout diagram and a Fast Mode speedometer overlay illustrating the June 1 2026 launch NEW
Models

Claude Opus 4.8 Adds Dynamic Workflows + Fast Mode

Anthropic ships Claude Opus 4.8 just 41 days after 4.7 with Dynamic Workflows for parallel subagents and a Fast Mode delivering up to 2.5x output tokens per second.

Aerial photograph of a hyperscale data-center campus at dusk, illustrating the infrastructure behind Google
Models

Google I/O 2026: Gemini 3.5 Flash and Spark Agent Land in Gemini App

Google at I/O 2026 on May 19 unveiled Gemini 3.5 Flash, a personal agent called Gemini Spark, and a new multimodal series called Gemini Omni. 3.5 Flash is now the default model in the Gemini app and A

Illustration of the Anthropic and SpaceX Colossus 1 partnership in Memphis showing a hyperscale data center hall packed with NVIDIA GPU racks dedicated to Claude AI workloads
Models

Anthropic and SpaceX Sign Compute Deal for 300MW Colossus 1 Capacity

On May 6, 2026, Anthropic agreed to take all compute capacity at SpaceX's Colossus 1 data center in Memphis — more than 300 megawatts and over 220,000 NVIDIA GPUs — and immediately raised Claude API a

OpenAI GPT-5.5 Instant model card inside ChatGPT showing the new default model with reduced hallucination on legal, medical, and financial queries while preserving low-latency Instant-tier responses
Models

OpenAI Releases GPT-5.5 Instant as New Default ChatGPT Model

OpenAI on May 5, 2026 began rolling out GPT-5.5 Instant as the new default ChatGPT model. It replaces GPT-5.3 Instant, cuts hallucination on legal, medical, and financial queries, and keeps the low-la

Microsoft Agent 365 admin dashboard showing agent inventory across Microsoft, AWS Bedrock, and Google Cloud, plus Defender and Intune controls governing locally running AI agents on Windows endpoints
Models

Microsoft Agent 365 Hits General Availability with Cross-Cloud Registry and Local Agent Controls

On May 1, 2026, Microsoft made Agent 365 generally available at $15 per user per month with a unified observe-govern-secure control plane, cross-cloud registry sync with AWS Bedrock and Google Cloud,

Anthropic Claude Security dashboard showing repository vulnerability scan results, severity rankings, and CSV export options on a developer workstation
Models

Anthropic Opens Claude Security to Enterprises in Public Beta for Vulnerability Scanning

On April 30, 2026, Anthropic released Claude Security as a global public beta for Claude Enterprise customers, adding scheduled scans, validated findings, and CSV/Markdown exports built on Opus 4.7.

Enterprise back-office workflow diagram with Salesforce Agentforce Operations agents handling invoice auditing, onboarding, and purchase order tasks across coordinated process steps
Models

Salesforce Launches Agentforce Operations to Automate Back-Office Bottlenecks

On April 29, 2026, Salesforce announced the general availability of Agentforce Operations — AI agents that complete back-office work end-to-end with 30+ pre-built blueprints and Instant Blueprints doc

Anthropic pricing page with Claude Code logo and dollar sign representing the A/B test removal of Claude Code from the 20 dollar monthly Pro subscription plan
Models

Anthropic Tests Pulling Claude Code From $20 Pro Plan — Infrastructure Costs Signal Pricing Shift Ahead

On April 21, 2026, Anthropic quietly updated its pricing pages and support documentation to remove Claude Code access from its $20/month Pro plan, listing Claud

Google Cloud Next 2026 conference stage with Gemini Enterprise Agent Platform logo and A2A protocol diagram representing the unified AI agent development platform announcement
Models

Google Cloud Next 2026: Vertex AI Reborn as Gemini Enterprise Agent Platform With A2A Protocol and 200+ Models

At Google Cloud Next 2026 on April 22, 2026, Google unveiled a sweeping reorganization of its enterprise AI portfolio under a single new brand: the Gemini Enter

Pentagon building with Google Gemini AI model visualization and classified data center concept representing military AI deployment negotiations after Anthropic phase-out
Models

Pentagon Moves Toward Google Gemini After Six-Month Anthropic Phase-Out Over Military AI Policy

Google is in active negotiations with the Pentagon to deploy its Gemini AI models inside classified Department of Defense systems, following the US military's d

OpenAI and Cerebras AI chip collaboration concept with wafer scale chip and data center infrastructure representing the 20 billion dollar three-year commitment
Models

OpenAI Commits $20B+ to Cerebras Chips Over Three Years, Receives Equity Stake Up to 10%

OpenAI has agreed to spend more than $20 billion over three years on servers powered by Cerebras chips — more than double the $10 billion deal the two companies

Anthropic Claude Opus 4.7 AI model abstract visualization showing high resolution vision capability and agentic task processing nodes
Models

Anthropic Launches Claude Opus 4.7 with High-Resolution Vision, Task Budgets, and Faster Agentic Performance

Anthropic launched Claude Opus 4.7 on April 16, 2026, releasing its most capable generally available model to date. The model introduces several new capabilitie

OpenAI GPT cybersecurity model abstract visualization showing restricted access shield with verified security researcher network and vulnerability detection
Models

OpenAI Unveils Restricted GPT-5.4-Cyber Security Model, Taking Different Approach Than Anthropic's Mythos

OpenAI unveiled GPT-5.4-Cyber on April 15, 2026 — a restricted-access AI model designed specifically for offensive and defensive cybersecurity work — following

Business performance bar chart with AI neural network overlay showing widening gap between AI leaders and laggards in corporate revenue growth
Models

PwC Study: Top 20% of Companies Capturing 75% of AI's Economic Gains

Three-quarters of AI's economic gains are being captured by just 20% of companies, according to PwC's 2026 AI Performance Study released April 13. The research

Google CEO Sundar Pichai speaking at a podium with AI circuit board graphics and the US Capitol representing the national AI leadership interview on 60 Minutes
Models

Google CEO Sundar Pichai: America Must Lead AI Development 'Boldly and Responsibly'

Google CEO Sundar Pichai called artificial intelligence "the most profound technology yet" and urged the United States to lead its development in a CBS 60 Minut

Elon Musk and Sam Altman silhouettes in a courtroom setting with OpenAI logo representing the high-stakes lawsuit over CEO removal and company control
Models

Musk vs. OpenAI Trial: Jury Selection Set for April 27 as CEO Ouster Bid Escalates

Elon Musk escalated his long-running lawsuit against OpenAI during the week of April 7-11, 2026, seeking not only damages but the explicit ouster of CEO Sam Alt

Anthropic Claude logo with CoreWeave data center infrastructure and revenue growth chart representing the 30 billion dollar annual run rate milestone
Models

Anthropic Hits $30 Billion Revenue Run Rate, Signs CoreWeave Infrastructure Deal

Anthropic disclosed on April 10, 2026 that its annual revenue run rate has surpassed $30 billion — a staggering leap from the $9 billion run rate the company re

Meta Muse Spark AI model launch graphic with Meta Superintelligence Labs branding and Llama open-source logo crossed out representing the proprietary shift
Models

Meta Launches Muse Spark, Its First Closed Proprietary AI Model, Abandoning Open-Source Llama Strategy

On April 8, 2026, Meta debuted Muse Spark — the first model from its new Meta Superintelligence Labs, headed by chief AI officer Alexandr Wang. The model is clo

OpenAI logo on a stock market ticker board with IPO documents and retail investor graphic representing the 852 billion dollar valuation
Models

OpenAI IPO Targets Retail Investors as $852 Billion Valuation Reshapes AI Finance

OpenAI CFO Sarah Friar told CNBC on April 8, 2026 that the company intends to allocate IPO shares to retail investors as it prepares for a public market debut,

Anthropic Mythos AI model abstract visualization with cybersecurity shield and code representing restricted Project Glasswing launch
Models

Anthropic Debuts Mythos, Its Most Capable Model Yet, Restricted to 12 Cybersecurity Partners Under Project Glasswing

On April 7, 2026, Anthropic unveiled a preview of Mythos — described internally as the company's most capable model to date — but deliberately restricted access

OpenAI Anthropic and Google logos united behind a shield blocking Chinese AI lab model distillation attempts via the Frontier Model Forum
Models

OpenAI, Anthropic, and Google Form Joint Front Against Chinese AI Model Distillation

OpenAI, Anthropic, and Google announced on April 6, 2026 — reported by Bloomberg — that they are sharing intelligence through the Frontier Model Forum to detect

Meta Llama 4 Scout and Maverick multimodal AI models launch graphic with 10 million token context window visualization
Models

Meta Releases Llama 4 with Scout and Maverick Models, Natively Multimodal and 10M Token Context

Meta launched Llama 4 Scout and Llama 4 Maverick on April 5, 2026, framing the release as a direct response to competitive pressure from Chinese open-source mod

US Pentagon building with Anthropic logo and courtroom scales of justice symbolizing the national security designation legal battle
Models

Anthropic Wins Court Injunction Against Pentagon's National Security Ban — DOJ Appeals

A federal judge issued an injunction on March 26, 2026 ordering the Trump administration to rescind its designation of Anthropic as a "Supply-Chain Risk to Nati

Google DeepMind Gemma 4 open-weight model release with Apache 2.0 license badge removing enterprise deployment restrictions
Models

Google DeepMind Launches Gemma 4 Under Apache 2.0 License, Removing Key Enterprise Barrier

Google DeepMind released Gemma 4 on April 2, 2026 under the Apache 2.0 license — a first for the Gemma model family. Prior Gemma releases were governed by Googl