GLM-5.1 Tops SWE-Bench Pro With Zero NVIDIA Hardware
Z.ai's GLM-5.1 scores 58.4 on SWE-bench Pro, edging out GPT-5.4 and Claude Opus 4.6, after being trained on 100,000 Huawei Ascend chips with no US silicon.
Articles Tagged "Benchmarks"
Claude Mythos Preview Finds Thousands of Zero-Days
Anthropic's restricted Claude Mythos Preview model autonomously discovered thousands of high-severity vulnerabilities across every major OS and browser, including bugs hiding in plain sight for 27 years.
AutoAgent Builds Its Own Harness, Tops Two Benchmarks
Kevin Gu's MIT-licensed AutoAgent lets a meta-agent engineer and hill-climb its own agent harness overnight, claiming the top GPT-5 slot on TerminalBench and first place on SpreadsheetBench.
Claude Sonnet 4.6 vs GPT-5.4: Same Price, Different Wins
Claude Sonnet 4.6 and GPT-5.4 cost nearly the same per token but win on opposite benchmarks. Here is where each model leads and which to pick for your workload.
Best AI Models for Code Generation - April 2026
Claude Opus 4.6 and GPT-5.4 lead different code benchmarks in April 2026 - pick based on your workflow, not one score.
Microsoft Open-Sources Harrier, a New Embedding Leader
Microsoft's Harrier-OSS-v1 family delivers three MIT-licensed multilingual embedding models, with the 27B variant claiming top spot on Multilingual MTEB v2 at 74.3.
Agent Consensus, Uncertainty Anatomy, and ARC-AGI-3
Three papers from today's arXiv: why multi-agent consensus is often a lottery, how to decompose LLM uncertainty into three actionable components, and what ARC-AGI-3 reveals about frontier AI's limits.
Better Planning, Faster Benchmarks, CFO Reality Check
Three new arXiv papers show how to build more reliable planning agents, cut benchmark costs by 70%, and why LLMs fail at long-horizon financial decision-making.
Best AI for Data Analysis - March 2026
Claude Opus 4.6 leads LiveSQLBench at 36.4% while ChatGPT's Code Interpreter dominates spreadsheet workflows - picking the right model depends on whether you need SQL, CSV analysis, or visualization.
Best AI for Web Browsing and Computer Use - 2026
GPT-5.4 leads OSWorld-Verified at 75.0% for desktop computer use while Claude Sonnet 4.6 matches human performance at 72.5% for half the price.
Best AI for Creative Writing - March 2026
Claude Opus 4.6 leads the Mazur Writing Benchmark at 8.56 while Claude Sonnet 4.6 tops EQ-Bench Creative Writing with 1936 Elo, making Anthropic the clear winner for fiction.
Best AI Models for Video Generation - March 2026
Seedance 2.0 leads the Artificial Analysis Elo rankings at 1,269, but Kling 3.0 is the most practical choice for global API access with native 4K at 75 cents per minute.
