frontier
65 stories and discussions about frontier, aggregated from every source we track.
Introducing the MiMo-V2.6 series: frontier intelligence, all the modalities, built in public.
Announcing Gemini 4 Argon, our frontier model for real-world coding, enterprise knowledge work, and cyber defense, rolling out soon.
A benchmark where frontier language models drive a real comma-equipped Toyota through a cone course, one command at a time, with a human supervisor ready to brake.
In one of my classes I asked the question I was afraid to ask but I just needed the answer to: “Who is afraid of not getting a job after graduating?” About eighty percent of the 150 people in the room…
Selling snake oil to the United States Congress is an ancient American craft, and the frontier artificial intelligence industry is currently attempting the most audacious hustle in modern corporate history.
阶跃星辰于2023年4月成立，以“智能阶跃，十倍每个人的可能”为使命。阶跃星辰坚定自研超级模型，积极布局算力、数据等关键资源，发挥算法和人才优势，已完成 Step-1 千亿参数语言大模型和 Step-1V 千亿多模态大模型的研发，在图像理解、多轮指令跟随、数学能力、逻辑推理、文本创作等方面性能达到业界领先水平。
From our March 2025 arXiv paper on RL conversion trajectories to building a sub-40ms...
I first played Prince of Persia in 1995 on an IBM PC XT. I still go back to it from time to time: the rotoscoped animation, the way the p...
Three robot policies given five instructions they should refuse. How often they refused, how often they carried them out.
HomeBody gives frontier vision-language models a humanoid embodiment through persistent spatial memory and composable skills, without environment-specific training or additional policy learning.
Evaluating frontier models on scientific work across chemistry, biology, and materials science, in a real research facility.
Google, OpenAI and Anthropic are closer to setting up a proposed independent body to develop standards for frontier artificial intelligence (AI), as debate...
turbopuffer is pushing the frontier of search. To do that, we have to fundamentally redesign our storage architecture so the vector index is no longer primary.
Open-source AI agent decision layer for Codex, Claude Code, Hermes, Antigravity and VS Code. TypeSafe Jev MCP routing, 38 recipes, local gates and receipts; optional Laya-MLX on Apple Silicon. - qu...
The $15.6 billion legal startup Harvey built its business around training AI models like OpenAI’s GPT-4 to do specialized work for lawyers. Recently, however, soaring artificial intelligence costs have nudged it to…
The Next Frontier: Welcoming AI Pioneer Jürgen Schmidhuber to Sakana AI
Strands harness is a fully assembled, customizable, state-of-the-art agent you run locally or deploy anywhere.
Jev isn't perfect. Here are some jagged edges we are aware of with jev-1.13. Many of these will be fixed in later versions.
GPT-6 Sol and Luna push the cost efficiency frontier by halving cost relative to GPT-5.6 Sol and Luna. Intelligence Index and Coding Agent Index scores remain level with GPT-5.6, with progress in some evaluations and…
Phylo cut inference cost 60% while doubling users month-on-month, running Biomni Lab's long-horizon biology agents on open models on Fireworks.
We’re partnering with Accenture on independent evaluation of frontier AI—part of our recent commitment to embed evaluators at Anthropic. Both we and Accenture expect to invest at least $1 billion to build capacity in…
turbopuffer is pushing the frontier of search. To do that, we have to fundamentally redesign our storage architecture so the vector index is no longer primary.
Cohere's Embed 5 delivers state-of-the-art retrieval across multimodal, multilingual, and financial documents - now available in Pro and Fast tiers.
Models are acting beyond intended limits, forcing an unprecedented pause on AI training.
Analysis of Xiaomi's MiMo-V2.6-Flash and comparison to other AI models across key metrics including quality, price, performance (tokens per second & time to first token), context window & more.
722+ open ML research and engineering roles at 28 frontier AI labs: OpenAI, Anthropic, xAI, Mistral AI, Cohere and more. From official career pages, refreshed every 6 hours.
With frontier AI companies’ money on the line, a mutual insurer would enforce standards, peer review incidents, and pool safety research and development.
A benchmark where frontier language models drive a real comma-equipped Toyota through a cone course, one command at a time, with a human supervisor ready to brake.
September 2026 Artificial intelligence is an extremely powerful technology. It has potential to improve lives, advance science and strengthen our economies. At the same time, the rapid development of frontier AI models…
Plus! Felonybench; Frontier Capabilities; SEO; Company Demographic Pyramids; Fighting for Inference; Diff Jobs
Today, the world can’t see what’s going on inside AI labs. Anthropic is proposing new metrics that would give the public visibility into frontier AI development.
A practitioner’s guide to frontier engineering. Ten principles for working with AI agents and shipping dramatically faster.
Frontier AI models were supposed to replace document processing. Instead, the category is growing as enterprises demand more precise, traceable, and reliable data extraction.
Panel Review makes four frontier models from OpenAI, Anthropic, Google, and xAI argue over your coding agent's riskiest changes before they execute. Only what survives the argument ships. Set up in under a minute with…
How I planned, built, and refined a mobile app with Claude Opus 5.5
The first frontier LLM with ten million tokens of context.
OpenAI has publicly detailed a more restrictive security approach for frontier reinforcement...
Reliability-constrained, multi-objective optimization for LLM and agent programs - obielin/reliopt
The rapid capability gains of frontier language models are widely attributed to improved reasoning abilities, yet this cannot be verified as raw CoT traces in closed-source systems are hidden. By registering a simple…
Spend tokens on judgment, not typing. An MCP-native autonomous SDLC pipeline: frontier models plan and review, local models implement — engineering discipline on a $20/month budget. - motock/fagan
Build once, add models as they launch, and choose where every job runs.
How CrowdStrike gave us meaningful advancements in cybersecurity-focused AI and why this has the ingredients to work at scale.
We tested 29 AI coding models on 60 real tasks and priced every run. One model reaches 93% of the top score for 2% of the cost. See the full data.
In most of the cases you don't need (Astra|Fable|Opus|%paste-a-new-LLM-name-released-this-week%).
GPT-6 Sol and Luna push the cost efficiency frontier by halving cost relative to GPT-5.6 Sol and Luna. Intelligence Index and Coding Agent Index scores remain level with GPT-5.6, with progress in some evaluations and…
Joint statement by international leaders on control of frontier AI models.
A task-level LLM router for pi: Jev classifies the task, a local selector picks a Pareto knee. Five tasks cost 65 percent less than fixed Sonnet.
Plus! Felonybench; Frontier Capabilities; SEO; Company Demographic Pyramids; Fighting for Inference; Diff Jobs
Strands harness is a fully assembled, customizable, state-of-the-art agent you run locally or deploy anywhere.