Magic

Building frontier code models to automate software engineering

No reviews yetby Magic AI, Inc.Pricing unavailable

Starting price

Pricing unavailable

Visit Magic Compare alternatives

Overview

Magic is an AI research and development company dedicated to building frontier code models designed to automate software engineering and research tasks. The organization combines frontier-scale pre-training, domain-specific reinforcement learning, ultra-long context windows reaching up to 100M tokens, and inference-time compute to create advanced code intelligence. Magic's overarching objective is to build safe artificial general intelligence (AGI) by automating AI research and code generation, aiming to solve alignment and capability challenges more reliably than human efforts alone. Backed by $515 million in funding and supported by thousands of GB200 GPUs in partnership with Google Cloud, Magic operates as a focused team of researchers and engineers addressing foundational problems in code synthesis and AI safety.

Best for: Software engineering teams and AI researchers seeking frontier code models and long-context capabilities.

Pros and cons

What works well

  • Pioneering research into ultra-long 100M token context windows.
  • Substantial infrastructure backing including thousands of GB200 GPUs.
  • Integrated approach emphasizing safety, alignment, and automated software engineering.

Where it falls short

  • Commercial product availability and pricing plans are not publicly listed.
  • No public API or platform integrations explicitly documented on the site.

Pricing

User reviews

Frequently asked questions

What does Magic do?

Magic is an AI company building frontier code models designed to automate software engineering and AI research.

What is Magic's context window capacity?

Magic has researched and developed context windows reaching up to 100 million tokens.

Who backs Magic?

Magic has raised $515 million from investors including Nat Friedman, Daniel Gross, CapitalG, Elad Gil, Sequoia, Jane Street, and Eric Schmidt.

What is Magic's approach to AI safety?

Magic follows an AGI Readiness Policy to evaluate, monitor, and mitigate existential risks associated with AI capabilities, aiming to use automated research to solve alignment reliably.

Magic alternatives