model labs & companies: 119 curated resources
model labs & companies
119 resources · page 4 of 4
Microsoft Research Blog
Microsoft Research blog with posts from Microsoft researchers on AI, systems and science.
Prime Intellect docs
Prime Intellect documentation for its compute marketplace, RL environments and training tools.
Reflection developer docs
Reflection AI developer docs for using its Beam model through the API, a coding agent or self-hosted deployment.
Sber on Habr
by Sber
Sber's company blog on Habr with Russian-language technical posts on GigaChat, GigaIDE and AI tooling.
SpaceXSI / Grok API docs
by xAI
xAI Grok API overview: models, quickstart and tools such as function calling, web search, code execution and remote MCP.
The "think" tool
by Anthropic
Anthropic post on the think tool, a scratchpad step that improves Claude's performance in complex tool use.
Tinker docs
Tinker overview documentation: training API for researchers to fine-tune open-weight language models.
Using GPT-6 (OpenAI API)
by OpenAI
OpenAI guide to using its GPT-6 model family for reasoning, coding and tool use, with migration notes.
Yandex on Habr
by Yandex
Yandex company blog on Habr with Russian-language engineering posts, including YandexGPT and AI code review.
August 2026 Risk Report (Anthropic)
by Anthropic
Anthropic August 2026 Risk Report published under its Responsible Scaling Policy.
DeepSeek-V3 Technical Report
by DeepSeek-AI
arXiv paper (2024) DeepSeek-V3 Technical Report: MLA attention, DeepSeekMoE and FP8 training.
Eval awareness in Claude Opus 4.6's BrowseComp performance
by Anthropic
Anthropic engineering post on Claude Opus 4.6 showing evaluation awareness on the BrowseComp benchmark.
GLM-4.5 ARC tech report
arXiv paper (2025) GLM-4.5 technical report on Z.ai's agentic, reasoning and coding foundation models.
Kimi K3: Open Frontier Intelligence
arXiv paper (2026) Kimi K3 technical report from Moonshot AI.
LoRA Without Regret
Thinking Machines Lab post LoRA Without Regret: when LoRA matches full fine-tuning, with hyperparameter guidance.
Luma Unified Models (Uni-1)
by Luma AI
Luma AI Uni-1 page: multimodal reasoning model that generates images.
MERA paper
by Alena Fenogenova et al. (SberDevices)
arXiv paper (2024) introducing MERA, a benchmark for evaluating LLMs in Russian.
MiniMax-M1 paper
by MiniMax
arXiv paper (2025) describing MiniMax-M1, scaling test-time compute with lightning attention.
MiniMax Sparse Attention
by Xunhao Lai et al. (Peking University)
arXiv paper (2026) describing MiniMax Sparse Attention, the attention design accompanying MiniMax-M3.
Muon is Scalable for LLM Training (Moonlight)
by Jingyuan Liu et al. (Moonshot AI)
arXiv paper (2025) by Moonshot AI showing the Muon optimizer scales to LLM training, with the Moonlight model.
CLOVA Tech Blog
NAVER CLOVA tech blog with technical posts on HyperCLOVA X models and their serving.
Nous Research blog
Nous Research blog with posts on Hermes models, the Psyche training network and research.
Qwen3 Technical Report
by An Yang et al. (Qwen Team, Alibaba)
arXiv paper (2025) Qwen3 Technical Report describing the Qwen3 model family, architecture and training.