1
Relent less AI self-evolution
AI自我进化新突破：将搜索代码从1260行压减至75行，准确率提升7.7%，token消耗减少4倍。
Article URL: https://github.com/001TMF/harness-forge Comments URL: https://news.ycombinator.com/item?id=48526233 Points: 1 # Comments: 0
2
I built a one click installer on token compression engine for AI tools
一个能一键安装 token 压缩引擎的开源工具，让 AI 工具更轻量高效。
the is nothing to explain about it , if I get some ups on my message I will make the tool free for all find it at https://slipstream.li Or skip the ma…
3
Show HN: Lore – LLM proxy for coding agent context and memory management
比压缩算法好2倍：Lore用智能记忆管理让AI编码代理告别68分钟/天的重复解释，总召回率提升至2.6倍。
Article URL: https://withlore.ai/ Comments URL: https://news.ycombinator.com/item?id=48464573 Points: 4 # Comments: 0
4
TokenTamer A proxy that reduces LLM token usage through context compression
实时压缩代码上下文，降低LLM API成本50-80%，可直接嵌入的代理工具。
Article URL: https://github.com/borhen68/TokenTamer Comments URL: https://news.ycombinator.com/item?id=48458633 Points: 1 # Comments: 1
5
Headroom compresses everything your AI agent reads before it reaches the LLM
AI代理读取内容前自动压缩，最高减少90% token消耗，实现成本断崖式下降。
Article URL: https://pypi.org/project/headroom-ai/ Comments URL: https://news.ycombinator.com/item?id=48349275 Points: 1 # Comments: 0
6
EarlyTom: Early Token Compression Completes Fast Video Understanding
CVPR 2026最新研究：通过早期Token压缩实现快速视频理解，颠覆传统视频处理效率瓶颈。
arXiv:2605.30010v1 Announce Type: new Abstract: Video large language models (Video-LLMs) have demonstrated strong capabilities in video understanding …
7
Accelerating Constrained Decoding with Token Space Compression
通过压缩词空间加速约束解码，让LLM输出精准服从语法规则，显著提升效率。
arXiv:2605.29986v1 Announce Type: new Abstract: To guarantee that an LLM's outputs conform to a specified structure, context-free grammar (CFG) decodi…
8
Show HN: Memory for LLM apps that cuts input tokens up to 80% (avg 68%)
GitHub开源项目，让LLM应用拥有长期记忆，同时将输入token平均削减68%，大幅降低API成本。
Article URL: https://github.com/Tem-Degu/streetai-memory Comments URL: https://news.ycombinator.com/item?id=48249509 Points: 1 # Comments: 0
9
LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMs
提出高效视觉编码器，解决Video LLM长视频中视觉token爆炸难题，突破帧扩展瓶颈。
arXiv:2605.17260v1 Announce Type: new Abstract: The fundamental challenge in scaling Video Large Language Models (Video LLMs) to long-form video lies …
10
Fre-Res: Frequency-Residual Video Token Compression for Efficient Video MLLMs
提出频域残差压缩方法，大幅减少视频MLLM的token数量，高效且不损失性能。
arXiv:2605.16366v1 Announce Type: new Abstract: Video MLLMs face a persistent tension between spatial fidelity and temporal coverage: preserving fine-…