Google DeepMind releases Gemma 4, its most capable open AI models yet

作者 PulseAugur 编辑部 · [3 个来源] · 2026-04-02 16:00

Google DeepMind has released Gemma 4, a new family of four open-source models ranging from 2 billion to 31 billion parameters. These models are designed for advanced reasoning and agentic workflows, with the 31B version achieving the third-highest rank on the Arena AI leaderboard and outperforming models 20 times its size. The smaller Gemma 4 models are optimized for on-device use and multimodal capabilities, supporting vision and audio processing with extended context windows. AI

影响 Sets a new benchmark for open-source model performance and efficiency, enabling advanced AI capabilities on local hardware and mobile devices.

排序理由 Release of new, highly capable open-source models from a major AI lab (Google DeepMind) with strong benchmark performance.

在 Google DeepMind 阅读 →

AI 生成摘要 · Google Gemini · 来自 3 个来源。我们如何撰写摘要 →

Google DeepMind releases Gemma 4, its most capable open AI models yet

报道来源 [3]

Google DeepMind TIER_1 English(EN) · 2026-04-02 16:00

Gemma 4: Byte for byte, the most capable open models

Gemma 4: Our most intelligent open models to date, purpose-built for advanced reasoning and agentic workflows.
dev.to — LLM tag TIER_1 English(EN) · Jubin Soni · 2026-05-15 02:36

Run Gemma 4 on Your Laptop — A Hands-On Guide to Google's Latest Open Multimodal LLM

<p>If you've been watching the open-source LLM space, you've probably noticed it's been a great couple of years. Llama, Mistral, Phi, Qwen — a whole zoo of models you can download and run on your own machine. Google's entry into that zoo is <strong>Gemma</strong>, and the fourth …
Mastodon — mastodon.social TIER_1 English(EN) · Techino · 2026-04-29 14:57

🧠 LLM UPDATE Google DeepMind just dropped Gemma 4—four open models, from 2B to 31B, with the 31B hitting #3 on Arena AI. The 31B outperforms models 20x its size

🧠 LLM UPDATE Google DeepMind just dropped Gemma 4—four open models, from 2B to 31B, with the 31B hitting #3 on Arena AI. The 31B outperforms models 20x its size. Smaller ones? Built for on-device and multimodal. https:// deepmind.google/technologies/g emma/ # Gemma # OpenSource #…

报道来源 [3]

Gemma 4: Byte for byte, the most capable open models

Run Gemma 4 on Your Laptop — A Hands-On Guide to Google's Latest Open Multimodal LLM

🧠 LLM UPDATE Google DeepMind just dropped Gemma 4—four open models, from 2B to 31B, with the 31B hitting #3 on Arena AI. The 31B outperforms models 20x its size

相关实体

相关话题