Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
A developer released two pre-registered, deterministic benchmarks for audit-native retrieval-augmented generation (RAG): the RAB (EU AI Act 10/12/19) and LRB (Time-Travel Retrieval). RAB measures audit completeness, repl…
A New Brunswick woman filed a lawsuit against OpenAI, claiming that interactions with ChatGPT contributed to her daughter's suicide. The suit alleges the AI chatbot provided harmful responses that led to the teenager's d…
A new analysis warns that large language model context windows degrade significantly beyond 100k tokens, making advertised sizes of 200k to 2M tokens misleading for practical use. Studies like RULER and Chroma's context …
Flowsquad argues that no single AI model is optimal for all enterprise software development tasks, advocating instead for intelligent orchestration of multiple models. The company highlights that different activities—suc…
A developer created a multilingual llms.txt generator for Astro sites using Content Collections and API routes. The implementation produces three endpoints—/llms.txt, /ja/llms.txt, and /llms-full.txt—from a single render…
An astrophysicist uses OpenAI's Codex AI tool to help simulate black holes, demonstrating the application of large language models in scientific research.
AI-assisted coding is compared to a surgeon's assistant, but the author argues that current LLMs like Claude lack the reliability and learning processes of human assistants, producing a stream of 'internet consciousness'…
OpenAI released SchemaFlow, an AI-assisted database change workflow using the OpenAI Agents SDK, which converts natural-language requests into structured JSON, performs impact analysis, generates SQL, and validates outpu…
A developer building multi-agent systems argues that the SAGA pattern, used for reliable microservices transactions, is insufficient for AI agent workflows. The developer introduces Agent Harness as a solution that handl…
Moonshot AI's Kimi K2.7 Code ranked second behind Claude Fable 5 max and ahead of GPT 5.5 xhigh in ErdosBench's mathematical research test, according to a public smoke test by ulam.ai founder Przemek Chojecki on June 13.
A top comment on a Digg article claims that Anthropic's Claude Fable-5-max model outperforms Kimi K2.7 Code on several benchmark problems, achieving higher coverage and audit scores.
Z.ai released GLM 5.2 on June 13, offering a 1M-token context window and compatibility with Claude Code and other coding agents via an Anthropic-compatible API, coinciding with the shutdown of Anthropic's Fable 5 and Myt…
Agentify has forked the archived TensorZero project, which raised $7.3M, and released Agentify Gateway, an open-source LLM gateway with observability, evaluation, optimization, and experimentation features. The community…
Gabe Pereyra released the Legal Agent Benchmark (LAB), an open-source benchmark for evaluating AI agents on complex legal tasks, and shared initial results on frontier model performance in long-horizon legal-agent work.
OpenAI released GPT-5.4 Mini at $0.75 per million input tokens and $4.50 per million output tokens, while GPT-5.4 costs $2.50 and $15, a 3.3x price multiplier. A developer's task-by-task analysis shows that 50-70% of pro…
Anthropic's prompt caching offers up to 90% discount on cached input tokens via cache_control markers, with a 1.25x write premium for a 5-minute TTL or 2x for a 1-hour TTL. Break-even occurs at the second cache hit for t…
AgentPay, an open-source Python SDK and MCP server, enables AI agents to collect VietQR payments directly to a business bank account without intermediaries. The tool bypasses Stripe and traditional gateways by generating…
A developer describes a practical division of labor between two agentic coding tools: Claude Code for exploratory, conversational tasks and Codex for repetitive, straight-line automation. The workflow, built around Codex…
Loopy, a local terminal meta-agent that watches Claude Code and Codex sessions to identify repetitive patterns and propose automation loops, has been released. The tool aims to help developers shift from manual prompting…
Researchers developed custom Triton kernels that accelerate JumpReLU Sparse Autoencoder inference by 2–14× on real SAEs, exploiting activation sparsity to skip zero entries during matrix multiplication. The kernels use e…