PulseAugur
EN
LIVE 05:03:12
日本語(JA) 約1兆パラメーターで画像・音声の理解やコーディングに対応、用途別に調整できるオープンウェイトAIモデル「Inkling」登場 https:// fed.brid.gy/r/https://gigazine .net/news/20260716-inkling/

Thinking Machines Lab releases Inkling, a 1T-parameter open-weight multimodal AI model

Thinking Machines Lab has released Inkling, an open-weight AI model with approximately 1 trillion parameters designed for fine-tuning. This model supports multimodal understanding of text, images, and audio, and allows users to adjust computational effort for different tasks. Inkling was trained on 45 trillion tokens and aims to serve as a foundation for organizations to build specialized AI applications by incorporating their unique knowledge. AI

IMPACT Inkling's open-weight nature and multimodal capabilities could accelerate the development of specialized AI applications across various industries.

RANK_REASON Frontier-lab model release with system card and open weights. [lever_c_demoted from frontier_release: ic=1 ai=1.0]

Read on Mastodon — mastodon.social →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Thinking Machines Lab releases Inkling, a 1T-parameter open-weight multimodal AI model

COVERAGE [1]

  1. Mastodon — mastodon.social TIER_1 日本語(JA) · [email protected] ·

    Introducing Inkling, an open-weight AI model with approximately 1 trillion parameters that supports image and audio understanding and coding, and can be adjusted for different use cases. https://fed.brid.gy/r/https://gigazine.net/news/20260716-inkling/

    約1兆パラメーターで画像・音声の理解やコーディングに対応、用途別に調整できるオープンウェイトAIモデル「Inkling」登場 https:// fed.brid.gy/r/https://gigazine .net/news/20260716-inkling/