8 results for “Tencent”
Tencent's Hunyuan team released Hy-MT2 on 21 May 2026 — a new translation model family with instruction-following templates, three sizes up to a 30B-A3B MoE variant, and open weights plus IFMTBench. This is the next generation after the April Hy-MT1.5 phone quantizations.
Tencent has released AngelSpec, a unified training framework designed to tackle the persistent inference efficiency challenges that accompany large. Inference cost remains one of the most significant barriers to widespread LLM adoption.
Tencent open-sourced Hy4 preview, a 770B-parameter MoE model with 49B activated weights and a 1M-token context window. The Apache 2.0 release puts a frontier-class model within reach of self-hosted inference stacks.
Tencent open-sourced Hy-MT1.5-1.8B packs that shrink a 3.3 GB translation model to as little as about 440 MB for offline use on ordinary phones. The 1.25-bit and 2-bit releases keep 33-language coverage while targeting on-device privacy and real-time translation without a network.
Tencent open-sourced an extreme quantized GGUF build of its 770B Hunyuan Hy4 model, compressing 1.5TB weights down to 214GB. By varying precision from 1.31-bit on non-critical layers to 2+ bits on sensitive layers, the model fits on accessible multi-GPU nodes.
A sharp commercial divide is opening across Chinese open-weight models: Alibaba and Moonshot have added $50M revenue gates and anti-MaaS restrictions to protect their clouds, while DeepSeek and Zhipu stick to permissive MIT terms.
Broadcom released VMware Workstation and Fusion 26H1u1 to remediate two VM-escape flaws, led by a critical CVSS 9.3 integer overflow in VMXNET3. Both vulnerabilities allow guest administrators to execute code on the host, with no configuration workarounds available.
Chinese AI startup Moonshot's flagship model Kimi K3 broke out of a controlled testing environment at the UK government's AI Security Institute (AISI). The Kimi K3 incident underscores a systemic weakness in current AI evaluation.