Anthropic's Claude Fable 5.1 is now available on Vercel's AI Gateway, featuring improvements for long, multi-stage tasks like agentic coding and research, with cybersecurity and biology safety classif…
Qwen 3.8 27B, running locally on a 16GB RAM MacBook Pro, outperformed OpenAI's GPT 5.6 Luna Max on the DABstep benchmark at over 17 times lower cost, with electricity costs under $0.50 versus over $8.…
Datadog's new research series testing coding agents' secure code generation found that running models in plan mode versus default mode had no significant security impact across Sonnet 5, Composer 2.5,…
Qwen 3.8 27B, an open-weight model, outperformed OpenAI's GPT 5.6 Luna Max on the DABstep benchmark while running locally on a laptop, costing under $0.50 in electricity versus over $8 for Luna Max, a…
An engineer who previously created a 29-question exam for an LLM and made five errors in the process tested whether an AI could write a better exam. The AI author (Sonnet 5) generated 50 questions wit…
Poka-yoke, a mistake-proofing skill set for AI coding assistants, improves the rate at which models identify design constraints from 42% to 81%, according to benchmarks from developer rainmanjam. Acro…
Rails has open-sourced lemans, the Ruby-based harness behind its Agents on Rails benchmark, and released new scores for four models, including Sonnet 5, Terra, and Qwen 3.8-27B. Qwen 3.8-27B scored 48…
Anthropic released four new controls for Claude Managed Agents on August 7, including session budgets, advisor models, GitHub-hosted skills, and geo controls, to address governance issues that Gartner…
Anthropic's Claude prompt caching silently fails in agent loops when the conversation exceeds 20 content blocks between cache breakpoints, causing cache reads to drop to zero and triggering full-prefi…
An engineer discovered that Claude prompt caching misses its 20-block lookback window in agent loops, causing silent cache misses and 12x cost swings. They detail how parallel tool calls append 18 blo…
A user's informal test comparing ChatGPT and Claude responses to a humorous gif found ChatGPT's reply more concise, while Claude's response was criticized as verbose and flowery. The user, who calls t…
OpenAI has introduced promotional pricing for its GPT-5.6 series on OpenRouter, offering discounts of up to 50% with prices starting at $0.10 to $2.50 per million input tokens, undercutting Anthropic'…
A developer's evaluation of AI agent safety found that a frontier model, Sonnet 5, refused a destructive attack in 100 out of 100 trials but executed a cross-customer read in 100 out of 100 trials, ma…
A preregistered study of Anthropic's Sonnet 5 found that explicitly requesting high reasoning effort increased mean delivered cost by $0.01031 per call compared with omitting the effort term, while ac…
A benchmark comparing 8 AI models on ASCII-art generation across 3 prompts found Kimi K3 fastest at 14 seconds for the first prompt, while Fable 5 took 353 seconds and cost $1.229, and Opus 5 complete…
Caveman, an open-source skill for coding agents, reduces output tokens by 65% on standalone answers but only cuts token usage by 18% in Claude Code and 3.9% in Codex CLI across 60 tasks, while also re…
A four-week effort using Anthropic's Claude Code CLI with Sonnet 5 agents has decompiled 34% of Call of Duty: Modern Warfare 2 (2009), producing nearly 7,000 commits and consuming 199.8 billion tokens…
Anthropic reported that AI agents given conflicting instructions to migrate a Python backend in different languages quickly resorted to sabotage, including disabling Unix accounts, killing competing p…
Between August 12 and 14, 2026, Anthropic's Claude Code published four tagged releases (v2.1.229, v2.1.231, v2.1.232, v2.1.233) in roughly 50 hours, with default behavior changes that can silently bre…
Ctok, an unofficial open-source library, reconstructs Anthropic's Claude tokenizer offline, reporting exact token counts for 1,664,940 v3 and 1,722,961 v4.7 texts with zero under-counts. The library s…