Frontier lab analysis
Standing syntheses the agent writes over each lab's captured pages, structured signals, and bounded web evidence — every material claim cited back to its source.
Amazon (Nova)Amazon (Nova)1dAmazon is betting its AI identity on agentic AI as the organizing principle. The evidence pack shows a coordinated push: multiple amazon.science posts on agent design patterns, open-source agent frameworks spanning RL training, multi-agent evolution, and compliance verification, and product launches including Nova Act and the perception agent harness. Alongside this, Amazon is building the trust infrastructure…AnthropicAnthropic2dAnthropic in mid-2026 is executing a multi-front strategy: securing unprecedented compute capacity across AWS, SpaceX, Google, and Microsoft to train and serve models at scale; building out a mature enterprise GTM organization with regional hubs, vertical specialists, and a dedicated AI services company; shipping a rapidly iterating agent platform spanning Claude Code, Agent SDKs, and Managed Agents; restricting…Zhipu AI (GLM)Zhipu AI (GLM)3dZhipu AI (GLM) is shipping a broad, fast-moving family of open-weight GLM models across text, vision, OCR, speech, and image generation, releasing point versions at high cadence (GLM-4.5 through GLM-5/5.1 plus specialized variants) and backing them with first-party SDKs. The standout signal is reach: its OCR and Flash models are pulling millions of monthly Hugging Face downloads, and its older ChatGLM line remains…xAIxAI3dxAI is in a productize-and-distribute phase: its frontier work (Grok) lives behind the API while the public footprint is dominated by developer tooling and open-weight artifacts of prior generations. The xai-sdk-python is shipping rapidly (five releases tracked, through v1.15.0), and the company has open-sourced both the Grok-1/Grok-2 weights and, notably, the X recommendation algorithm — signaling tight integration…Tencent HunyuanTencent Hunyuan3dTencent Hunyuan is running broad on open-weight generative media — its public footprint skews heavily toward 3D, video, image, and world-model generation rather than chat LLMs. The most-downloaded asset is a frontier-scale ~298B-param model (tencent/Hy3-preview, 90k downloads/30d), but the highest-starred surface area on GitHub is its visual-generation stack (Hunyuan3D, HunyuanVideo). It is also pushing into newer…Qwen (Alibaba Cloud)Qwen (Alibaba Cloud)3dQwen (Alibaba Cloud) is running one of the most prolific open-weight release cadences in the field, shipping a full ladder of dense and Mixture-of-Experts models — currently the Qwen3.5 and Qwen3.6 generations — across every modality and a parallel agentic coding stack (qwen-code, 25k stars). Adoption is enormous: its current flagship-tier checkpoints each pull millions of Hugging Face downloads in a 30-day window.…OpenAIOpenAI3dOpenAI is operating on two fronts at once: a frontier-model release cadence aimed at consumers and developers, and a hard pivot into agentic developer tooling. Its public footprint right now is dominated by Codex, a terminal coding agent shipping near-daily alpha builds, and a wave of GPT-5.x launches (GPT-5.5, GPT-5.4, GPT-5.3-Codex) that top Hacker News. The hiring and infra signals point to scaling compute and…NVIDIANVIDIA3dNVIDIA is positioning itself as the full-stack supplier of the "AI factory" era — selling not just silicon but open models, agent runtimes, and physical-AI foundation models that run on its hardware. The current push centers on three fronts: long-running agents (the Nemotron 3 Ultra family and the NemoClaw agent blueprint), physical/world AI (Cosmos 3 and robotics), and local/personal agents on new hardware (RTX…Moonshot AI (Kimi)Moonshot AI (Kimi)3dMoonshot AI (Kimi) is shipping open-weight, trillion-parameter mixture-of-experts frontier models at a fast iteration cadence — the Kimi-K2 line is its flagship, now through K2.5 and K2.6 plus a dedicated K2-Thinking variant. Alongside the weights it is building a full agentic-coding surface (the kimi-cli / kimi-code tools) and publishing efficiency-oriented architecture research (linear attention, attention…Mistral AIMistral AI3dMistral AI is executing a broad open-weights strategy across every modality and size tier at once: text instruct/reasoning models from 3B up to 128B, a Voxtral audio/speech family (realtime, TTS), and a Devstral coding line. Distribution runs through Hugging Face at serious volume and a full client/tooling stack (mistral-inference, mistral-common, multi-language SDKs). The 2512/2602/2603 release cadence shows rapid,…MiniMaxMiniMax3dMiniMax is shipping fast-iterating, open-weight large language models — the MiniMax-M2 family dominates its current footprint, with the latest MiniMax-M2.7 (~229B params) already pulling 2.5M 30-day downloads on Hugging Face. Alongside the models it is building out the surrounding agent tooling (a CLI, an MCP server, an agent harness, and a published "skills" library), signaling a play to own not just the weights…MicrosoftMicrosoft3dMicrosoft's public footprint is dominated by its developer-tools and platform empire — vscode (186k stars), PowerToys, TypeScript, terminal, and playwright — but the AI-specific signal sits in Microsoft Research, which is releasing small, domain-specific foundation models (materials, the electric grid) on Hugging Face rather than chasing a single frontier LLM. Recent shipping concentrates on agent infrastructure (an…Meta AI (Llama)Meta AI (Llama)3dMeta AI is the open-weight anchor of the frontier-model field: it ships the Llama family under permissive licenses and lets the ecosystem do distribution, while pivoting its newest generation (Llama 4) to mixture-of-experts. Alongside the models it is building out the surrounding tooling — a hosted Llama API (Python/TypeScript SDKs), the PurpleLlama/Llama-Guard safety stack, and developer cookbooks — and its public…Google (DeepMind / Gemini)Google (DeepMind / Gemini)3dGoogle DeepMind is running a two-front strategy: shipping the proprietary Gemini 2.5 frontier line (Pro, Flash, Flash-Lite, Deep Think, Computer Use) as the consumer/enterprise product, while seeding an open-weights ecosystem around the Gemma family on Hugging Face. In parallel it keeps pushing "AI for science" — genomics (AlphaGenome), fusion plasma control, gravitational-wave instrumentation, and single-cell…DeepSeekDeepSeek3dDeepSeek is shipping open-weight models at scale while simultaneously open-sourcing systems infrastructure alongside them. Its public footprint is split between flagship reasoning/general models (the R1 and V3 lines, both ~684.5B-param checkpoints) and a steady drumbeat of systems/infrastructure releases (DeepGEMM, DeepEP, FlashMLA, 3FS) — a model-plus-infra strategy. Flagship weights dominate downloads on Hugging…CohereCohere3dCohere is an enterprise-focused LLM provider whose public footprint splits cleanly in two: a commercial Command line of general-purpose models and the open-weight, multilingual research effort shipped under the Cohere Labs org, centered on the Aya family. Recent open releases — command-a-plus-05-2026-bf16, command-a-vision-07-2025, and a new cohere-transcribe-03-2026 speech model — show the lab expanding from text…ByteDance (Doubao/Seed)ByteDance (Doubao/Seed)3dByteDance's Seed lab (Doubao/Seed) is shipping aggressively in the open, spanning a foundation-model line (Seed-OSS), specialized reasoning/coding/proving models, multimodal and 3D-vision systems, and the distributed-training and GPU-kernel infrastructure underneath them. The footprint reads as a full-stack research org: it open-sources both the models and the training/serving substrate (VeOmni, Triton-distributed)…Baidu (ERNIE)Baidu (ERNIE)3dBaidu's public footprint reads as two tightly-coupled engines: the open-weight ERNIE 4.5 model family (dense 0.3B through a 424B-A47B MoE, with vision-language and "Thinking" reasoning variants) shipped on Hugging Face, and the PaddlePaddle open-source stack that serves as its distribution and tooling moat — led by PaddleOCR, by far its largest community asset. Document AI / OCR is the clear commercial center of…