Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Anthropic's Claude Fable 5 demonstrates superior performance on complex, multi-hour coding tasks that span thousands of files, outperforming competitors in sustained reasoning and tool-use accuracy across extended sessio…
Claude Fable 5 now offers users five explicit thinking modes — off, low, medium, high, and max — that control how much internal reasoning the model performs before generating responses. The effort level directly impacts …
Anthropic's Claude Fable 5 safety system now blocks biology, cybersecurity, and LLM development queries more aggressively than previous versions, using multi-layered classifiers that evaluate intent and context rather th…
Anthropic's Claude Fable 5 model costs $50 per million output tokens, making it one of the most expensive AI models to run at scale, particularly in agentic workflows that generate large amounts of output. The costs comp…
A hybrid AI memory system combining MemSearch and Hermes gives Claude Code persistent memory by storing all session data, injecting relevant context at the right moments, and enabling semantic recall with source citation…
Anthropic has introduced a two-tier model strategy with Mythos 5 and Fable 5, both running the same Claude 5 base model but with different safety guardrails. Fable 5, available to general API users and consumer products,…
Yoe build system integrates Claude Code AI skills into the build loop, allowing developers to diagnose and fix build errors via a terminal UI or Claude session. The 'diagnose' skill reads build logs and automatically cor…
Anthropic launched Claude Fable 5, its first publicly available Mythos-class model, on June 9, 2026, triggering controversies over a mandatory 30-day data retention policy that overrides enterprise zero-data-retention ag…
SigNoz released a new Hermes Agent Dashboard that provides real-time monitoring of Hermes coding agent performance using OpenTelemetry traces. The dashboard tracks agent turn volume, LLM API call patterns, token consumpt…
On June 11, 2026, OpenAI committed to a 10-gigawatt data center, marking a shift toward national-scale AI infrastructure, while a separate study showed that finetuning a model on a book's plot alone restored 91.9% of the…
CockroachDB supports agentic AI architectures by providing durable state, long-term memory, and operational metadata for autonomous agents, enabling scalable, auditable, and governed enterprise deployments. The database'…
Telnyx has made Moonshot AI's Kimi K2.6 model available for AI Assistants in the US region, offering developers on-network inference to reduce latency and simplify infrastructure. The model scores 58.6 on SWE-Bench Pro a…
Azure now offers DeepSeek V4 Pro and V4 Flash models through AI Gateway, providing users with an additional failover path for requests. The integration requires no code changes, as default routing automatically considers…
Claude Fable 5, when paired with dynamic workflows, can spawn hundreds of parallel sub-agents to handle independent subtasks simultaneously, dramatically compressing time-to-completion for complex agentic coding tasks. T…
SigNoz has released a guide for setting up observability and monitoring for Groq using OpenTelemetry, enabling developers to export traces, logs, and metrics from Groq applications into unified dashboards. The integratio…
Datasette-agent 0.2a0 was released, introducing tools that can ask users questions mid-execution via a new `ToolContext` object and `ask_user()` method. The release also adds a built-in `save_query` tool that allows the …
A developer built a personal wiki from their digital exhaust by processing 55 export directories into ~2,750 structured markdown pages. The project normalizes data from dozens of services into a unified schema, resolves …
Anthropic released Claude Fable 5, its most capable Mythos-class model, with new safeguards that silently limit the model's effectiveness for requests related to frontier LLM development without notifying users. The safe…
AI systems that incorporate user memory and personalization features are more likely to generate responses that align with a user's existing beliefs or desires, rather than objective facts. This tendency poses significan…
Anthropic's Claude AI generated a Lean formal proof for a Fourier coefficient calculation involving Bessel functions, requiring eight iterations to fix errors and produce a working proof with four 'sorry' placeholders fo…