Grok 2.5

xAI's open-weight release of their 2024 flagship model — 270B dense transformer, competitive with GPT-4 at release, requires 8x 40GB GPUs

View API Documentation

Model Specifications

ArchitectureTEXT

Parameters270B

Familygrok

VRAM (Q4)API Only

Released Aug 2025 under Grok 2 Community License (restricts use for training other AI models). ~500GB total. SGLang recommended.

#xai#flagship#reasoning#open-weightSource

Share this Model

Send this model's specs directly to your community.

Post

Similar Models

Grok-1

314B

xAI massive open model

Grok 3

314B

xAI's open-weight dense base model at 314B parameters, released under Apache 2.0. Designed for research and large-scale fine-tuning; not instruction-tuned by default.

Llama 3.3 70B

70.55B

Refined Llama 3 with superior following

Related Guides

How much VRAM do you really need?

A complete breakdown of quantization levels and VRAM overhead for running local models.

Best GPUs for Machine Learning in 2026

Comparing NVIDIA and AMD options for the best speed-to-dollar ratio.

GGUF vs EXL2 vs AWQ

Understanding local AI formats and which one to pick for your specific hardware.