Skip to content

Grok-2

Chat, code and reason with Grok-2

Visit official site

What Grok-2 is

The model is described as xAI's frontier language model, released with Grok-2 mini to Grok users on the X platform. The material states an early version was tested on the LMSYS leaderboard as sus-column-r and scored above Claude 3.5 Sonnet and GPT-4-Turbo in overall Elo.

What you can do with it

  • Chat, write code and reason with the model
  • Follow instructions with attention to accurate facts
  • Reason over retrieved content and use tools
  • Handle visual maths reasoning and document-based questions

Application scenarios

  • Answer complex programming questions with Grok-2
  • Use Grok-2 mini where a fast response is needed
  • Collect insights and data during academic research

Getting started

  • Open the X platform and update the X app to test Grok-2
  • Reach both models through the Grok tab in the X app
  • Pick Grok-2 or Grok-2 mini for the interaction
  • Use advanced functions such as visual understanding or coding tasks

Pros & cons

✓ What we like

  • Retrieved-content reasoning and tool use are stated to be improved
  • Visual maths reasoning and document-based questions are covered

! What to watch out for

  • Access is described through the X app and a Grok tab
  • The LMSYS result is reported for an early version named sus-column-r

FAQ

Which models were released?

The material names Grok-2 and Grok-2 mini, both released to Grok users on the X platform.

What did the LMSYS test show?

The material states an early version tested as sus-column-r scored above Claude 3.5 Sonnet and GPT-4-Turbo in overall Elo.

Last reviewed: 2026-11-15

More AI chatbot tools

View all →

How we review