Skip to content

AI News

New models, tools worth trying, and enough context to understand what's happening in AI.

29Sep2026
Models & ArchitectureNew

Introducing Claude Sonnet 5.5: 30% Faster with Coding Leap at $2

Claude Sonnet 5.5 delivers 30% faster speed, scores 70.6% on Terminal-Bench 4.0, and cuts effective task cost by 30% at unchanged $2/$10 pricing.

  • 23Sep2026
    Models & Architecture

    Claude Opus 5.5: Frontier Agentic Coding at 40% Lower Cost

    Claude Opus 5.5 delivers frontier agentic coding and a 1M context window at 40% lower cost than Opus 5, matching Fable 5.1 for enterprise workloads.

  • 26Jul2026
    Models & Architecture

    Claude Opus 5: Stronger Than Fable 5, at Half the Price

    Claude Opus 5 delivers superior agentic performance and a 1M context window at half the cost of Fable 5, setting a new baseline for enterprise AI workflows.

  • 20Jul2026
    Models & Architecture

    Kimi K3: The Open Frontier Intelligence

    Kimi K3 is Moonshot AI's 2.8T-parameter open-weight model that rivals GPT-5.6 and Claude Fable 5 in coding, with a 1M-token context window.

  • 12Jul2026
    Models & Architecture

    OpenAI Releases GPT-5.6

    OpenAI releases GPT-5.6 with Sol, Terra, and Luna tiers, featuring advanced reasoning modes, 750 tps Cerebras support, and Zero Data Retention APIs.

  • 10Jun2026
    Models & Architecture

    Claude Fable 5: Anthropic's New Flagship AI Model

    Claude Fable 5 is Anthropic's flagship Mythos-class AI model, with a 1-million-token context window and agentic autonomy for long, complex tasks.

  • 03Jun2026
    Tools & SDKs

    Claude Code Introduces Dynamic Workflows for Parallel Agents

    Optimize large-scale engineering with dynamic workflows in Claude Code, using JavaScript orchestration to manage parallel agents and durable plans.

  • 02Jun2026
    Models & Architecture

    Claude Opus 4.8: 4x fewer code-flaw misses and parallel agents

    Claude Opus 4.8 keeps $5/$25 pricing and 1M context but is 4x less likely to miss its own code flaws, and adds Dynamic Workflows for parallel agents.

  • 28May2026
    Models & Architecture

    Claude Opus 4.7: stronger coding and the new xhigh effort level

    Claude Opus 4.7 provides enhanced software engineering autonomy, high-resolution vision, and a new xhigh effort level for advanced professional workflows.