Tagged with

3 articles found

K2 Horizon Just Redefined What ‘Open Source AI’ Actually Means, And It’s About Time

IFM’s K2 Horizon fleet drops six fully open models from 0.9B to 375B with complete training code, data, and checkpoints. Here’s why this changes everything.

#AI Benchmarks#Frontier Models#K2 Horizon...

AI Benchmarks

DeepSeek’s Dark Magic: Revolutionary AI Efficiency or State-Subsidized Dumping?

DeepSeek’s API costs 10-107x less than GPT-5.5 while matching its benchmarks. Is this Mixture-of-Experts genius, cache-hit wizardry, or something more politically complicated?

#AI Benchmarks#AI pricing#API economics...

AI Benchmarks

GLM 4.7’s Benchmarks Skyrocket, But Its Censorship and Real-World Gaps Raise Red Flags

While GLM 4.7 ranks #2 on Website Arena with impressive scores, users report increased censorship and skepticism about benchmark validity versus actual performance in coding and creative tasks.

#AI Benchmarks#censorship#GLM 4.7...