Large Language Models News
Sakana Fugu turns model orchestration into one API
Sakana AI is positioning Fugu as a single API that dynamically coordinates expert models for coding, reasoning, and other complex multi-step tasks.
By The AI Feed Desk
OpenAI brings ChatGPT and Codex to Samsung Electronics employees
OpenAI says Samsung Electronics is deploying ChatGPT Enterprise and Codex to all employees in Korea and all Device eXperience employees worldwide.
By The AI Feed Desk
OpenAI puts ChatGPT Enterprise spend into the admin console
OpenAI is adding credit usage analytics and updated spend controls for ChatGPT Enterprise, including ChatGPT and Codex usage by user, product, and model.
By The AI Feed Desk
MosaicLeaks makes research-agent privacy a query-log problem
ServiceNow researchers introduce MosaicLeaks, a benchmark showing how deep research agents can leak private enterprise facts through ordinary-looking web queries.
By The AI Feed Desk
DeepSWE makes coding-agent rankings a cost question
DeepSWE's June 20 leaderboard update separates frontier coding agents by pass rate, cost, output tokens, and agent steps across long-horizon software tasks.
By The AI Feed Desk
MLPerf Mobile v6.0 gives on-device LLMs a real test surface
MLCommons added standardized Android LLM tests to MLPerf Mobile v6.0, including Llama 3.2 1B, Llama 3.2 3B, and Llama 3.1 8B Instruct workloads.
By The AI Feed Desk
MAI-Code-1-Flash moves across GitHub Copilot before enterprise access
GitHub says Microsoft's small coding model is expanding across Copilot CLI, app, chat, IDE, mobile, and Xcode surfaces before Business and Enterprise rollout.
By The AI Feed Desk
GPT-5.5 Instant makes health a default ChatGPT test
OpenAI says GPT-5.5 Instant improves ChatGPT health responses for free users, with physician rubrics, HealthBench evaluations, and production factuality monitoring.
By The AI Feed Desk
OpenAI's rare-disease study makes old genome cases worth reopening
OpenAI says o3 Deep Research helped experts reanalyze 376 previously unsolved rare-disease cases and establish 18 diagnoses after clinical review.
By The AI Feed Desk
Z.ai releases GLM-5.2 for long-horizon coding work
Z.ai's GLM-5.2 pairs a 1-million-token context pitch with long-horizon coding benchmarks, public docs, API pricing, and an MIT-licensed Hugging Face model card.
By The AI Feed Desk
Gemini API adds TTS streaming as media model shutdown dates arrive
Google's Gemini API changelog added streaming speech generation for a preview TTS model and set near-term shutdown dates for older Imagen and Veo model IDs.
By The AI Feed Desk
OpenAI shows GPT-5.4 improving a medicinal-chemistry reaction
OpenAI says GPT-5.4, connected to Molecule.one's Maria lab, found an additive that improved a difficult Chan-Lam coupling result across thousands of physical experiments.
By The AI Feed Desk
OpenAI uses deployment simulation to test models before release
OpenAI says replaying realistic conversation contexts helped forecast undesired behavior across GPT-5-series Thinking deployments before models reached users.
By The AI Feed Desk
Anthropic suspends Claude Fable 5 and Mythos 5 after US directive
Anthropic says it disabled Claude Fable 5 and Claude Mythos 5 for all customers after a US export-control directive covering foreign-national access.
By The AI Feed Desk
Google releases DiffusionGemma for faster local text generation
Google's DiffusionGemma is an experimental open text-diffusion model that generates blocks of text in parallel for lower-latency local workflows.
By The AI Feed Desk
Google brings Gemini models to Apple developers
Google says Apple developers can call Gemini models through Apple's Foundation Models framework and use Gemini inside Xcode.
By The AI Feed Desk
Anthropic releases Claude Fable 5 and Claude Mythos 5
Anthropic's first broadly available Mythos-class model arrives as Claude Fable 5, with sensitive requests routed to Opus 4.8 and Mythos 5 reserved for trusted access.
By The AI Feed Desk