AI Models

Claude Opus 4.7 Review: Same Price, 10 Points Smarter at Coding, and 24% Faster in Production

Anthropic's April 16 release of Claude Opus 4.7 reclaims the coding crown with 87.6% on SWE-bench Verified, writes its own tests before finishing a task, and — per Box's data — cuts API calls by 56%. Here's what actually changed and whether you should migrate.

Qwen 3.6 Plus Review: A Free 1M-Token Model That Actually Works

Alibaba's Qwen 3.6 Plus: 1M token context, free preview, real coding and multilingual chops. Where it replaces Claude and GPT — and where it clearly doesn't.

Claude vs GPT vs Gemini: 90 Days of Real Work in 2026

90 days running Claude, GPT, and Gemini side-by-side on production tasks. Benchmarks lie, real workflows don't. Here's which model wins which job in 2026.

MiniMax M2.7 Review: Claude Opus Performance at 1/17 the Price

MiniMax M2.7 hits frontier coding benchmarks at $0.30 per million tokens — 17× cheaper than Claude Opus. The self-evolving training trick and where it wins.

Gemma 4 on a MacBook: The Local LLM That Ships Real Code

Gemma 4 runs on 4GB, fits on a MacBook, and writes production code. Open weights, the exact setup, the benchmarks, and the honest limits — after two weeks of use.

Anthropic Just Built an AI Too Dangerous to Release — What Claude Mythos Actually Does

Anthropic built Claude Mythos, decided not to release it, and told the world why. Inside Project Glasswing — and what withholding a frontier model actually signals.

Anthropicが「危険すぎて公開できない」AIを作った——Claude Mythosの正体と、Project Glasswingの全貌

Anthropicが公開を見送った未公開モデル「Claude Mythos」と、背景にあるProject Glasswing。日本語で読める最長級の解説と、AI業界の構造分析。