DeepSeek is a Chinese AI research laboratory that has developed highly capable open-source language models including DeepSeek-V3 and DeepSeek-R1, notable for their efficiency and performance.
Open-weight models handled 78.4% of token volume through Vercel AI Gateway versus 21.6% for closed models in a September 19th daily snapshot, according to Vercel founder and CEO Guillermo Rauch. Verce…
A developer who maintains Alca, a provider-tracking site, documented four distinct categories of "free" AI model quotas after repeatedly exhausting paid tiers: harness-locked pools usable only through…
A developer detailed how frontier LLM development is shifting from pre-training scaling to test-time compute scaling, outlining three architectural regimes: sequential chain-of-thought expansion, leaf…
GreyNoise Intelligence reported on September 9 that a single attacker using hundreds of AI agents compromised 440 PaperCut NG/MF servers across 395 organizations in 48 countries on August 31, with 11 …
Open-weight models reached 56% of all token volume on Vercel's AI Gateway as of August 2026, up from 11% in April, according to Vercel AI Gateway data, marking the first time open models have overtake…
A custom 32-bit CPU gate-level crackme called HellGates, published by its creator on crackmes.one in November, was solved by GPT-6 in under 20-30 minutes in September 2026 on SRE-Bench, after going un…
A developer published a configuration guide for setting up OpenCode V2 with a single custom provider, Relax AI (Beta), authenticated via a saved account rather than environment variables or shell API …
Xiaomi's MiMo team launched a public dashboard at mimo.xiaomi.com/rl on September 16 that streams live training telemetry for two trillion-parameter models, MiMo-V2.6-Pro and MiMo-V2.6-Flash, with the…
Hacker News user Argonautlabs posted an Ask HN thread titled "DeepSeek v4.1 Flash on Local hardware, what tok/s do you see?" asking the community to report local inference token-per-second rates for D…
The LectuLibre team built a token-based chunking pipeline for translating full-length books with large language models, using Python, FastAPI, and Anthropic's Claude alongside DeepSeek for simpler pas…
Meta plans to deploy its third-generation custom AI processor, MTIA 450, code-named Arke, in data centers in the first half of 2027, targeting lower cost per dollar and per watt for AI inference, acco…
A builder documented a dual-boot Ubuntu 24.04/Windows workstation running four NVIDIA RTX6000 Blackwell Pro Max-Q GPUs (96 GB each, 384 GB total) with an AMD Ryzen Threadripper PRO 7985WX and 512 GB o…
A developer released WTAgent, a free, MIT-licensed CLI that turns ChatGPT Web and Claude Web accounts into Codex-style local coding agents. The tool drives a dedicated Chrome profile signed into the u…
A user reports that attempts to download DeepSeek v4.1 from Hugging Face repositories consistently return a 400 error, while downloads of other models work normally. The user says the failure is limit…
US export-control restrictions imposed on Anthropic and OpenAI have pushed cybersecurity professionals toward less-restricted Chinese AI models, with Hugging Face turning to Zhipu AI's GLM-5.2 in July…
Moonshot AI's Kimi K3 is now available on Amazon Bedrock, with AWS stating the model is the first open model to reach 2.8 trillion parameters and the first open-weight model on Bedrock to support expl…
DeepSeek's V4.1-Flash, a 552-billion-parameter MoE model, cuts KV cache memory from 389,120 bytes per token in the company's first 2023 model to 890 bytes, a 437-fold reduction, through techniques inc…
SemiAnalysis reported that DeepSeek's Engram architecture, which extends standard token embeddings with learned multi-token lookups, can lower required HBM capacity for models at the same quality by o…
A developer behind the open-source pi-box project demonstrated a pi agent running inside an AgentVM virtual machine autonomously installing and launching the game Doom, with the Freedoom splash screen…
DeepSeek v2 reduces memory footprint by 437x compared with DeepSeek v1, according to a video report. The source provides no further details on the model, its release date, or the benchmark behind the …