DeepSeek
DeepSeek is a Chinese AI lab, backed by the High-Flyer hedge fund, that shook the industry with open reasoning models trained at a fraction of Western costs. Its R1 release in early 2025 wiped hundreds of billions off US tech stocks and made training efficiency a global priority. All our DeepSeek coverage is collected here.
AdaPRA Lifts an 80B Qwen by 24.5 Points on Kaggle-Style Leaderboards
Carnegie Mellon researchers gave a data science agent a supervisor with adaptive checklists. With five attempts, an 80B Qwen gained 24.5 percentile points on…
DeepSeek open-sources its DeepInfer inference engine, and analysts calculate an 82% margin
DeepSeek has released its internal DeepInfer framework and the deepseek-recipe component under Apache 2.0. Third-party estimates put the gross margin of its AI…
DeepSeek releases V4.1-Flash-1017 and opens its Search API at $0.75 per 1000 calls
DeepSeek's updated Flash model, V4.1-Flash-1017, finishes tasks faster, writes shorter answers and is open source. Search and Fetch APIs cost $0.75 per 1000 calls.
Design Anything Keeps the Brief Intact in Large Projects: +25% Across 24 Artifacts
Synchronous MoE decoding makes GPUs wait for the busiest group
Conditional Program shrinks DeepSeek's n-gram memory from 203 to 51 GB at 2 bits
Vera Rubin NVL72 beats GB300 by up to 5.7x in tokens per megawatt in AA-AgentPerf
Apple's LensVLM compresses long agent context into images and teaches the model to "zoom"
Apple's LensVLM turns long-running agent context into images, and the model zooms into the parts it needs instead of relying on a lossy summary.
A honeypot repo with 1,774 stars: two filters for detecting malware and hallucinations
A Telegram bot sent a repo with 1,774 stars — a disguised malware trap. After auditing 321 repos from trending feeds, the author built two filters: one to detect…
200 OK, but the client gets nothing: three n8n agent bugs from production
Qwen 3.8 Max and DeepSeek V4 Compared on a Single Cross-Domain Question in PlyLoom
DeepSeek to lead seed round for robotics startup Winverse AI at a $400 million valuation
ByteDance Researchers Find a Weak Spot in DeepSeek's Memory-Saving Approach
MagicBeanStalk lifts gpt-oss-120b from 61.2 to 69.8 points on math
Authors show that 88% of tasks in a large RLVR corpus give no learning signal, and propose MagicBeanStalk: task synthesis added +8.6 points for gpt-oss-120b.
FrontierCode: Top AI Models Fail Maintainer Review in 42% of Attempts
Cognition and Artificial Analysis launched FrontierCode 1.1: Claude Opus 5.5 leads with 52% and a 58% pass rate, at $10.15 per task. GPT-6.1 Sol scores 49% for…