Tag

#AI benchmarks

4 articles

GPT-5.5 tops benchmarks but still hallucinates frequently at a 20 percent higher API cost

GPT-5.5 tops AI benchmarks but still hallucinates frequently, and its API cost has risen by 20%.

Alibaba's open model Qwen3.6 leads Google's Gemma 4 across agentic coding benchmarks

Alibaba's Qwen3.6 outperforms Google's Gemma 4 in agentic coding benchmarks, showcasing the power of efficient AI architectures.

Apr 1729

AI benchmarks systematically ignore how humans disagree, Google study finds

This article explains how human disagreement in AI benchmarking can lead to unreliable performance metrics and why current practices need to evolve to account for annotation variability.

Apr 464

Cohere releases open source model that tops speech recognition benchmarks

Cohere has released an open-source speech recognition model that outperforms industry leader OpenAI's Whisper in benchmark tests.

Mar 2772