Together AI expanded its Together Fine-Tuning service to support 17 new open-weight models, including GLM 5.3, DeepSeek-V4-Flash, Kimi K2.7-Code, Qwen 3.5 and Gemma 4, alongside live experiment tracki…
Databricks released Adaptive Instructed-Retriever, a retrieval model that dynamically adjusts how many sequential search steps it performs based on query difficulty, letting developers cap the maximum…
DeepSeek AI released DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts model with 552B backbone parameters, 196B additional Engram parameters, and a 1M-token context window, cutting global KV cache…
Researchers introduced KC-Bench, a dynamic interactive benchmark with 238 manually screened tasks for evaluating how LLM agents handle knowledge conflicts across world-knowledge, input inconsistencies…
DeepSeek-AI released the DeepSeek-V4 series on April 24, 2026, including the MIT-licensed DeepSeek-V4-Flash model with 284B total parameters (13B activated) and a one-million-token context window, and…
Alibaba released a preview of Qwen3.8-Flash-Next, a 125-billion-parameter model that activates only 6 billion parameters per token, achieving training costs roughly one-ninth of typical models of its …
Alibaba's Qwen team released Qwen3.8-Flash-Next on 26 August, a mixture-of-experts model that activates only 6 billion of its 125 billion parameters per token and is billed as an architecture preview …
Researchers from UC Berkeley and UT Austin released FreeToken, an edge-native Mixture-of-Experts serving engine that runs the 753B-parameter GLM-5.2 model on a single workstation GPU, achieving 35B at…
FreeToken, an open-source inference engine developed by UC Berkeley researchers including Shuo Yang and Xiaoze Fan, reports decoding DeepSeek-V4-Flash, a 284B-parameter Mixture-of-Experts model, at 22…
StateM, a stateful control system for long-horizon agents, ranked #1 on Hugging Face Daily Papers on 2026-08-18 and released a runbook and reproducibility package for DeepSeek-V4-Flash, achieving 88.8…
FreeToken, a new inference engine from FlashML, lets users run frontier models on gaming GPUs at interactive speeds, with Qwen3.6 35B running on an 8GB RTX 4060 laptop at 39 tokens per second, DeepSee…
FlashML released FreeToken, an edge-native Mixture-of-Experts (MoE) serving engine that runs 290B+ parameter frontier MoE models locally on consumer gaming PCs at interactive speeds. The engine suppor…
DeepSeek released DeepSeek-V4-Flash-Vision-Exp, an experimental multimodal model on its API platform on August 21, adding image understanding to its V4-Flash line while retaining text capabilities. Th…
A developer built a C99 inference engine that runs DeepSeek-V4-Flash, a 284B-parameter mixture-of-experts model, on a laptop with just 3.2GB of RAM by streaming weights off NVMe and caching only the e…
Telnyx has added DeepSeek-V4-Flash and Qwen3.8-27B to its Inference API, offering cost-efficient and low-latency options for high-volume and classification tasks. DeepSeek-V4-Flash is priced at $0.13 …
DeepSeek released DeepSeek-V4-Flash-Vision-Exp, an experimental multimodal model now live on the DeepSeek API Platform, matching DeepSeek-V4-Flash on text capabilities while making a major leap on mul…
A new benchmark, StateMemBench, with 234 multi-session scenarios, shows that LLM-based agent memory systems fail to track evolving state, and the proposed StateMem method improves current-state accura…
UC Berkeley's CyberGym benchmark shows AI agents can now reproduce real-world software vulnerabilities with 93.2% accuracy, up from 10-30% a year ago. The leaderboard leader, a Sangfor AI Agent runnin…
DeepSeek will raise peak-hour prices for its flagship V4 models by more than four times from current levels starting Aug 16, with the DeepSeek-V4-Flash model priced at US$1.32 per million output token…
Autonomous AI agents built on open-source frameworks breached Taiwanese government systems, compromised credentials, and probed a nuclear safety agency in a multi-day cyberattack, according to researc…