Skip to content

Reka Core

A multimodal LLM for image, video and audio context

Visit official site

What Reka Core is

The model is described as a GPT-4-level multimodal large language model with strong context understanding of images, video and audio. The material also presents it as one of two commercial comprehensive multimodal solutions on the market.

What you can do with it

  • Understand context across images, video and audio
  • Work with a 128K context window
  • Apply reasoning that includes language and mathematics
  • Use enhanced coding and agent workflows with support for 32 languages

Application scenarios

  • Image recognition and analysis
  • Video content understanding and editing
  • Natural language processing and generation

Model notes

  • The material describes it as a GPT-4-level multimodal LLM
  • It is described as one of two commercial comprehensive multimodal solutions
  • The card names multimodal understanding, reasoning, coding, agent workflows, multilingual support and deployment flexibility as strong areas

Pros & cons

✓ What we like

  • Reasoning covers language and mathematics
  • Coding and agent workflows are enhanced
  • Deployment flexibility is listed among its strong areas

! What to watch out for

  • Context understanding is described for image, video and audio inputs
  • The material names a 128K context window for the model

FAQ

How large is the context window?

The feature list states a 128K context window.

How many languages are supported?

The feature list states support for 32 languages.

Last reviewed: 2026-11-15

More AI coding tools tools

View all →

How we review