Technology ❯ Artificial Intelligence ❯ Model Evaluation ❯ Benchmarking

Performance Metrics

OpenAI Builds GPT-Red to Hunt Prompt‑Injection Flaws

OpenAI says its internal attacker discovered new exploit techniques that led to fixes cutting prompt‑injection failures in GPT‑5.6 Sol.

Chinese AI Firms Slash Training and Inference Costs With Sparse MoE and H800 Workarounds

Z.ai Releases GLM-5.2 Open-Weight Model Trained on Huawei Chips

Companies Slash AI Use After 'Token Shock' Drives Soaring Bills

AI Competition Shifts to Chips and Power as Anthropic Nears $1 Trillion and Dell Posts Blockbuster Quarter

DeepSeek Debuts V4 Preview With 1M‑Token Context, Tuned for Huawei Chips

Google DeepMind Releases Gemma 4, Open AI Models Built for Phones to Workstations

OpenAI Rolls Out GPT-5.2 in Three Tiers for ChatGPT and the API

OpenAI Says Incentives Drive AI Hallucinations, Calls for Scoreboard Overhaul

September Research Blitz Reveals LLM Vulnerabilities and Rolls Out Real-World Benchmarks

OpenAI Launches GPT-5, Starts 60-Day Transition From Older Models

OpenAI Unveils First Open-Weight Models Since GPT-2, Run on Common Hardware

DeepSeek Upgrades R1 AI Model to Challenge Top Western Rivals