DeepSeek launches low-cost coding model to challenge Claude

A new coding model from DeepSeek matches top benchmark scores while cutting API costs sharply.

DeepSeekReported60 sec read

Laptop screen displaying code and data charts.

Aug 10: DeepSeek launches low-cost coding model to challenge Claude

Photo: Daniil Komov · Unsplash

What happened

DeepSeek released a new AI model targeted at coding tasks. Reports indicate the model costs 99 percent less to run than Anthropic's Claude. It scores within two benchmark points of Claude's performance.

The context

High API fees remain a key barrier for software teams running continuous AI code assistants. Ultra-low pricing from competitive Chinese labs keeps driving down market rates for developer tools.

Sources

WION ↗ via Google News Reported

See DeepSeek's Vikshy Score → · DeepSeek's models & pricing →