微软被曝收紧Token使用管理，将监控用量并把内部默认工作负载逐步切换至GPT-5.6 Sol以降本。导火索是泄露表格显示，部分员工为抬高数据排名大量调用AI，单人在28天内产生约2.8万美元Token成本，引发高层担忧；各部门差异明显，CoreAI最“烧钱”。
阿里千问发布多模态MoE模型Qwen3.8-Flash，并开源下一代架构Qwen3.8-Flash-Next（Qwen4雏形）。模型总参数125B，搭载51B N-gram Embedding，每token仅激活6B；原生支持26万token，可扩展至100万。训练成本为前代1/9，API定价输入每百万Tokens 1元、输出3元。
AI服务商Writer推出旗舰模型Palmyra X6，基于开源模型GLM-5.2后训练变体，并升级智能代理运行框架，可将企业基础任务的Token成本降低高达50%，旨在破解AI“越用越贵”困境。
DeepSeek-V4-Pro-0813在硅基流动Day-0首发，主打1M上下文、低/高/最大三档推理强度，重心转向编码、工具调用与智能体工作流，维持MIT开源。定价为输入每百万token 1.32美元、输出3.96美元，缓存命中仅0.44美元，缓存成本优势显著。
Openai
$2.8
Input tokens/M
$11.2
Output tokens/M
1k
Context Length
Google
$0.49
$2.1
Xai
$1.4
$3.5
2k
$7.7
$30.8
200
Anthropic
$105
$525
$0.7
$7
$35
$17.5
$21
Alibaba
-
$4
$16
$2
$20
$6
$24
256
Bytedance
$0.8
128
$0.15
$1.5
Baidu
32
OneTool是一个统一的MCP服务器，通过Python API暴露100多种工具，大幅减少AI调用工具时的token消耗和成本，避免上下文污染，支持Web搜索、数据库、文件操作、图表绘制等多种功能。
基于cursor-agent CLI的MCP服务器，提供代码分析、搜索、规划和编辑等工具，帮助Claude Code减少token使用和成本
%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
Join Now
