Microsoft's new MAI Code 1.1 Flash gets crushed by Deepseek on both price and performance
Microsoft's new code model loses on both speed and cost to DeepSeek's cheaper alternative—a sign the price-performance bar just moved.
Why it matters
Code model capability and pricing are shifting rapidly; practitioners evaluating GitHub Copilot alternatives now have hard benchmark data showing cheaper, faster options exist. This is a genuine competitive moment in the frontier race.
The key facts
7 to knowMicrosoft MAI Code 1.1 Flash: 25% more token-efficient than predecessor, quarter the cost
DeepSeek V4 Flash outperforms MAI Code 1.1 Flash on benchmarks
DeepSeek V4 Flash is cheaper than Microsoft's offering
Competitive positioning: open AI narrative vs proprietary model strategy
MAI Code 1.1 Flash: 25% more token-efficient than predecessor, 1/4 the cost of prior version
Model comparison: proprietary vs. open-weight coding model capability race
Competitive pressure: frontier labs releasing cheaper, faster alternatives
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: Microsoft has released MAI Code 1.1 Flash, a code model for GitHub Copilot that's said to be 25 percent more token-efficient at a quarter of the cost of its predecessor. In benchmarks, though, it gets crushed by the cheaper Deepseek V4 Flash. The move fits a pattern: Microsoft talks up open AI,…