Let’s start with Grok 3 Mini. When we set out to build a fast, affordable mini model, we knew it would be good but even we didn’t expect it to be this good. Some highlights: – Grok 3 Mini tops the leaderboards on graduate-level STEM, math, and coding, outcompeting flagship
GENERATIVE AI
-

Grok 3 Demonstrates Expert-Level Enterprise AI Capabilities
By
–
Grok 3 excels at tough enterprise tasks that require specialized, expert-level knowledge. Customers across finance, healthcare, and legal have told us they’re seeing emergent capabilities they haven’t seen with any other models.
-

Grok 3 Family API Launch: Cost-Efficient AI Models
By
–
Meet the Grok 3 family, now on our API! Grok 3 Mini outperforms reasoning models at 5x lower cost, redefining cost-efficient intelligence. Grok 3, the world's strongest non-reasoning model, excels in tasks that need real world knowledge like law, finance, and healthcare.
-
Grok 3 and Grok 3 Mini now available on Poe platform
By
–
You can try Grok 3 at https://
poe.com/Grok-3, Grok 3 Mini at https://
poe.com/Grok-3-Mini, and on all platforms. (3/3) -
Grok 3 Mini: Lightweight reasoning for everyday questions
By
–
Grok 3 Mini offers lightweight reasoning capabilities and excels at clear, concise reasoning for everyday questions and logical problems. It’s designed for straightforward interactions where extensive expertise isn't needed. (2/3)
-

xAI Launches Grok 3 and Grok 3 Mini Models
By
–
Now available: Grok 3 and Grok 3 Mini! Grok 3 is xAI's flagship model with deep expertise across specialized domains. It gives in-depth answers to complex queries, drawing on its extensive knowledge across specialized fields. (1/3)
-

DeepSeek MoE Multi-Node Deployment Improves Throughput Latency
By
–
Our AI team just published a blog detailing how we've achieved improved throughput and latency with multi-node deployment of DeepSeek MoE models. Read more:
-
Master AI Systems: Prompt Engineering and Practical Applications
By
–
Immerse yourself in how Gemini and these AI systems work, spend every waking hour working with them so it's like second nature, become a ninja at prompting and system instructing, and use it from everything from deep research to vibe coding…
-
Building Tetris with VibeCoding: A Quick Example
By
–
Petit exemple de création du jeu Tetris en "VibeCoding" pic.twitter.com/WUpANd0H9b
— VISION IA (@vision_ia) 18 avril 2025Petit exemple de création du jeu Tetris en "VibeCoding"
-

Test-Time Scaling in Large Language Models: Survey
By
–
What, How, Where, and How Well? A Survey on Test-Time Scaling in Large Language Models
Paper: https://
arxiv.org/pdf/2503.24235
Repo: https://
github.com/testtimescalin
g/testtimescaling.github.io/
…
