Anthropic released Claude Sonnet 5.5 on September 28, 2026: same listed price as Sonnet 5, over 30% faster, up to 30% cheaper per task, and second worldwide on the Intelligence Index behind Opus 5.5.
Anthropic made Claude Code cloud sessions generally available with promo credits: $100 for Pro, $250 for Max. Eligibility rules, claim deadline, and all four ways to launch a session.
Claude Opus 5.5 opens the Anthropic 5.5 family: Fable 5.1-level performance at 40% lower running cost, $4/$20 per million tokens, and 30% faster output than Opus 5.
Grok 4.7 is SpaceXAI's most capable model for coding and knowledge work, launched September 21, 2026 at $2 input / $6 output per million tokens — same price as Grok 4.6, with major gains.
Meta's Muse Spark 1.3 lands in Muse Code and the Meta Model API: 25% fewer tokens, stronger prompt-injection resistance, and an agent that asks instead of assuming.
OpenAI and AWS brought the GPT-5.6 family (Sol, Terra, Luna) to the Kiro IDE with roughly 82% cost reduction — full pricing breakdown and how Kiro compares to Cursor and Copilot.
Grok 4.6 hits frontier level: 61 on the AA Intelligence Index, 69.9% on CursorBench, at $2/$6 per million tokens. The complete guide to performance, access, and rivals.
Cursor enters code hosting with Origin: repositories, pull requests, two-way GitHub sync, and Vercel, Depot and Buildkite integrations inside the AI editor, in an early paid beta.
Z.ai releases GLM-5.3 (Aug 14, 2026): beats Claude Opus 4.8 on Z.ai Code Bench with far fewer tokens, tops CyberGym at 84.5%, and open weights arrive in two weeks. Full guide.
أعلنت SpaceXAI عن إطلاق Grok 4.5، وهو نموذج من فئة Opus صُمم خصيصاً للبرمجة، والمهام المستقلة، والأعمال المعرفية المعقدة بفضل كفاءة استهلاك الموارد ومستوى الذكاء الفائق والتكامل مع Cursor.
Start typing to search
