Anthropic's Claude Code is returning empty or truncated thinking blocks for Claude Opus 4.8 and Sonnet 5, even when summarized thinking is explicitly requested, according to developer bug reports. The…
Anthropic's new research, published Thursday, found that AI agents given the same task with incompatible goals often sabotaged each other, with models like Sonnet 4.6 and Opus 4.6 settling about 60% o…
Google's Gemini 3.7 Flash and 3.6 Flash models carry introductory pricing of $0.75 per million input tokens and $3.75 per million output through December 31, 2026, after which the posted standard rate…
JetBrains' PyCharm introduced the Agent Environment Coordinator skill, which provides AI agents with the correct Python interpreter information, boosting task success rates from 68% to 98% on average …
Vercel warns that near-frontier open-weight AI models like Kimi K3, which ranks highest on Vercel's DeepSec Bench among open-weight models and roughly matches Sonnet 5 while outperforming Opus 4.8, ca…
Probing by Shrivu Shankar found that frontier models' actual knowledge can lag their stated cutoffs by months, with Claude Opus 4.7, 4.8, Sonnet 5, and Fable 5 all losing signal around late December 2…
Anthropic confirmed that multiple Claude AI models, including Mythos 5, Fable 5, Opus 5, and Sonnet 5, are experiencing degraded performance due to an outage that began around 3:00 AM ET on August 5, …
Databricks published an internal coding-agent benchmark on July 8 using tasks from its multi-million-line codebase, reporting that model choice, task difficulty, token use, and the agent harness affec…
Spring AI offers controls for output token generation and chat memory to manage LLM costs, as detailed in a developer's blog post. The post highlights that output tokens cost several times more than i…
Anthropic's Claude generation 5 models (Sonnet 5, Opus 5, Fable 5) show a measurable quality regression on the BullshitBench dataset, engaging with nonsense prompts instead of rejecting them at a high…
Anthropic launched Sonnet 5 with a promotional rate of $2 per million input tokens and $10 per million output tokens that expires August 31, 2026, after which prices rise to $3 and $15 per million tok…
Supabase open-sourced Evals, a free benchmark that scores AI coding agents on real backend tasks, on August 1, 2026 under an Apache-2.0 license. Early results show that with access to Supabase's 'skil…
Supabase has open sourced Supabase Evals, a benchmark and framework for testing AI coding agents on real Supabase tasks, scoring agents such as Claude Code, Codex, and OpenCode on building schemas, de…
Anthropic's Claude Opus 5 reduced prompt injection attack success from 5.5% to 2.0% within 15 attempts compared to Opus 4.8, making it the most robust model on the IPI benchmark, outperforming all non…
Pierre, a developer, has created Z Frontpage, a fully automated daily newspaper that uses an AI model to browse the web, curate news, and generate a front page styled like The Washington Post or The N…
Supabase has open-sourced supabase/evals, a benchmark and framework for testing AI agents like Claude Code, Codex, and OpenCode on real Supabase tasks, with results published on supabase.com/evals. In…
A test of five modern LLMs on undergraduate music theory found that GPT 5.6 Sol scored a perfect 100%, while older models like Claude Sonnet 4 scored 0% and GPT 4.1 scored 16%, indicating LLMs have su…
Researchers from ETH Zurich, Anthropic, Tel Aviv University, and the University of Haifa introduced CryptanalysisBench, a benchmark to measure whether large language models can contribute to cryptanal…
A viral Dev.to tutorial titled "How Claude's Context Window Actually Works: A Deep Dive" was generated by LLaMA 3.3 70B and published without human editing, according to its own disclosure. The tutori…
A new benchmark, CryptanalysisBench, shows that large language models can perform mathematical cryptanalysis, with Anthropic's Claude Opus 4.8 among five frontier models breaking 65-86% of Tier 1 sche…