Chinese content分析 3 万场 Arena 对战后，Claude Fable 5 的概念相似度排名出人意料
朋友分析 3 万场对战数据，发现 Claude Fable 5 的概念相似度排名和预期不同，挺有意思的。
Latest
Every published item, in order · 331 items
Sep 12
Sep 11
Chinese contentAI时代工程师技能变化：OpenAI GPT-6、Anthropic Claude Fable 5.1等新模型发布
朋友，OpenAI和Anthropic都发布了新模型，GPT-6和Claude Fable 5.1，还有Google等公司的语音转文本模型，错误率都低于4%，这些新模型和技能变化信息很实用。
Cursor releases private benchmark CursorBench 4.0
For developers, this matters because it provides a more realistic and difficult benchmark to test AI coding assistants, moving beyond public benchmarks that can be gamed.
Chinese contentClaude Fable 5.1 和 GPT-6 Astra 让专有模型与开源模型在网页开发上的差距重新扩大
Claude和GPT-6的新版本让专有模型在网页开发上又领先了，想了解最新差距的可以看看。
Sep 10
Sep 9
Chinese contentOpenAI发布GPT-6 Astra重塑Agent Arena前沿
OpenAI新模型GPT-6 Astra在性价比上表现突出，比Claude Fable便宜7%但差距很小。
Sep 7
GPT-6 Astra vs Claude Fable 5.1
Astra offers better value for financial tasks while Fable delivers superior structured documents, helping users choose the right model for specific workloads.
Sep 6
Chinese contentClaude Fable 5.1 (Max)登顶Agent Arena榜首
Anthropic的Claude Fable 5.1 (Max)在Agent Arena排名第一，性能领先但成本较高，用户满意度极高。
Chinese contentClaude Fable 5.1登顶Agent Arena排行榜
Anthropic的Claude Fable 5.1在真实任务测试中表现优异，性价比高，用户满意度遥遥领先。
Claude Fable 5.1 tops Agent Arena
Claude Fable 5.1 outperforms competitors in agent tasks, making it a strong choice for developers building autonomous systems.
Chinese contentGPT-6 Astra Max登顶Code Arena编程榜单
OpenAI新模型GPT-6 Astra Max在编程评测中夺冠，比第二名高35分，价格与Claude持平。
Sep 5
Chinese contentGPT-6 Astra与Claude Fable 5.1对比评测
OpenAI的GPT-6 Astra与Anthropic的Claude Fable 5.1正面PK，成本、编程能力实测对比
Sep 4
GPT-6 Astra Benchmarks Show Mixed Results
The ARC-AGI-3 performance suggests AGI may arrive sooner than anticipated, affecting global AI development timelines and investment strategies.
Chinese contentGPT-6 Astra与Claude Fable 5.1对比评测
OpenAI的GPT-6 Astra与Claude Fable 5.1的3D生成能力对比，详细展示了不同推理级别的影响。
Chinese contentMuse Spark 1.3重回WebDev编程榜前十
Meta的Muse Spark 1.3在编程榜单上跃升20多位，性能接近Claude和GPT-5.6，工具调用减少20%。
Chinese contentAnthropic推出Claude Fable 5.1限时测试
Anthropic新推Claude Fable 5.1，限时在Arena直接测试，号称编程和知识工作最强模型。
Sep 3
Claude Fable 5.1 decodes royalist message
Claude Fable 5.1 shows practical problem-solving abilities beyond standard benchmarks, useful for researchers working with historical texts.
Chinese contentAnthropic发布Claude Fable 5.1
Anthropic刚发布的Claude Fable 5.1能一次性生成游戏和3D世界，看看这10个实际案例有多惊艳。
Anthropic releases Claude Fable 5.1 and Mythos 5.1
Models cost 25% less with improved coding performance, making them competitive alternatives to OpenAI's offerings for developers in the region.
Chinese contentMeta 发布 Muse Spark 1.3 模型
Meta 新出的 Muse Spark 1.3 编码能力超 GPT-5.6，工具调用减少 20%，还支持长线程多任务处理。
Chinese contentProximal发布FrontierSWE v2编程基准
Proximal发布了新版编程基准FrontierSWE v2，Claude Fable 5.1在测试中表现最佳。
Sep 2
Claude Fable 5.1 SVG pelican
The $3.30 cost for high-quality SVG generation may interest designers and developers working with vector graphics.
Chinese contentClaude发布Fable 5.1与Mythos 5.1双模型
Anthropic发布Fable 5.1和Mythos 5.1，前者编程能力大幅提升且成本降低25%，后者专注网络安全和生命科学领域。
Claude Fable 5.1 animated pelican test
Fable 5.1's performance on scientific benchmarks and varying reasoning levels offers developers in ASEAN a cost-effective alternative to GPT-5.6 Sol.
Claude Fable 5.1 Released
Claude Fable 5.1 offers enhanced performance for knowledge work at no mentioned price change, making it a compelling upgrade for existing users.
Chinese contentAnthropic 发布 Claude Fable 5.1 和 Mythos 5.1
Anthropic 发布两款新模型，编程和科学能力大幅提升，价格更实惠，安全防护也增强了。
Chinese contentAnthropic发布Claude Fable 5.1和Mythos 5.1
Anthropic推出Claude 5.1双版本，科学基准分翻倍，缓存成本降75%，API有重大变更
Chinese contentAnthropic发布Claude Fable 5.1模型
Anthropic新发布的Fable 5.1模型在科学基准测试中翻倍得分，编码能力提升30%，同时成本降低45%。
Chinese contentAnthropic发布Claude Fable 5.1和Mythos 5.1
Anthropic发布Fable 5.1和Mythos 5.1，科研能力翻倍，缓存成本降75%，企业数据留存问题解决。
Chinese contentClaude Fable 5.1 登陆 OpenRouter
Anthropic 发布了 Fable 5.1，在编程和金融分析方面表现更好，OpenRouter 已可使用。
Chinese contentAnthropic发布Claude Fable 5.1和Claude Mythos 5.1
Anthropic新推Claude Fable和Mythos 5.1，专为编程和知识工作设计，比前代更强。
Aug 29
Zhipu releases GLM-5.3 model weights
GLM-5.3 offers advanced AI capabilities for coding and cybersecurity at lower cost than closed-source alternatives, with commercial use permitted under specific terms.