AI News
Tencent’s Hy4 Flash Packs 220B Parameters but Wakes Only 12B per Token
November 18, 2026
Tencent-Hunyuan’s Hy4 flash holds 220B parameters but activates only 12B per token. The repo claims big per-task savings over its own models, and those numbers still need outside checking.
Read moreRipple-65B Preview Claims 80.52 on TerminalBench 2.1 With a 65B Dense Model
November 17, 2026
Ricursive Intelligence says its 65B dense Ripple-65B Preview scores 80.52 on TerminalBench 2.1, ahead of far larger open-weight models. The scores are self-reported.
Read moreKV1-1.2B Is Trained for a 2-Bit KV Cache, and the Memory Math Is Striking
November 16, 2026
Klein ML’s KV1-1.2B was trained for a 2-bit KV cache, shrinking it to 60 MiB at 32K context while keeping 75.8% of fp16 quality. The catches are real.
Read moreAnalysis

A 9,500-Token Chat Bills 59,000. Here’s Where the Rest Goes
November 18, 2026

The Contract Fit. The AI Still Cited the Wrong Clause.
October 19, 2026

AI Model Selection Framework: Which Frontier Model for Which Job
April 29, 2026
Guides

‘You Are a Helpful Assistant’ Does Almost Nothing. Try This
November 16, 2026
Most system prompts describe a mood and change almost nothing. This step-by-step method turns vague intentions into checkable rules, examples and tests that visibly change what the model writes.
Your Docs Chatbot Cited the Handbook and Still Got It Wrong
October 23, 2026
One Timestamp Turned a $422 Bot Bill Into $1,863
October 21, 2026






