Pith Wave Signal

4 stories tagged #SWE-Bench Pro

Newest Most read
  1. Meta to Launch Advanced AI Coding Model to Rival OpenAI and Anthropic
    tech

    Meta to Launch Advanced AI Coding Model to Rival OpenAI and Anthropic

    Meta plans to release a significantly upgraded Muse Spark AI model with enhanced coding capabilities, challenging GPT-5.5 and Claude Opus 4.8 in key benchmarks.

    last wk. 1 min read
  2. Anthropic's Claude Sonnet 5 Challenges Opus Model at a Lower Price Point
    tech

    Anthropic's Claude Sonnet 5 Challenges Opus Model at a Lower Price Point

    Anthropic's new Claude Sonnet 5 model matches the performance of its top-tier Opus 4.8 on key benchmarks, offering developers near-equivalent intelligence at a significantly reduced cost.

    last wk. 2 min read
  3. Anthropic's Claude Fable 5 Shatters AI Benchmarks, Surpassing OpenAI's Leading Model
    tech

    Anthropic's Claude Fable 5 Shatters AI Benchmarks, Surpassing OpenAI's Leading Model

    Anthropic's new 'Mythos-class' Claude Fable 5 achieves a 161 on the Epoch Capabilities Index and dominates in software engineering tests, signaling a major shift in the enterprise AI landscape.

    last mo. 1 min read
  4. Kimi WebBridge Lets AI Agents Drive Your Browser Locally
    tech

    Kimi WebBridge Lets AI Agents Drive Your Browser Locally

    Moonshot AI launches Kimi WebBridge, a browser extension for local AI agent automation, keeping data private.

    2mo ago 1 min read