Anthropic just released Claude Fable 5 a Mythos-class model for general use, with safety classifiers that fall back to Opus 4.8 on ~5% of sessions
English summary
Anthropic released Claude Fable 5, their most capable publicly available model, and Claude Mythos 5 restricted to cyberdefense partners. Fable 5 demonstrated remarkable performance: it migrated a 50-million-line Ruby codebase in a day, beat Pokémon FireRed using raw screenshots, and scored highest on FrontierCode eval. Mythos 5 autonomously conducted genomics research across 138 species, outperforming a published Science paper with a 100x smaller model. The safety approach uses classifiers that silently fall back to Opus 4.8 on sensitive queries, with zero universal jailbreaks found in over 1,000 hours of testing. Pricing is $10/M input and $50/M output tokens, with free access through June 22 for limited plans.
Chinese summary
Anthropic 发布了 Claude Fable 5，这是他们公开可用的最强模型，以及仅限于网络防御合作伙伴的 Claude Mythos 5。Fable 5 展示了卓越性能：一天内迁移了 5000 万行 Ruby 代码库，仅使用原始截图击败了《宝可梦 火红》，并在 FrontierCode 评估中得分最高。Mythos 5 自主进行了跨 138 个物种的基因组研究，以缩小 100 倍的模型超越了已发表的《科学》论文。安全方法采用分类器，在敏感查询时静默回退到 Opus 4.8，在超过 1000 小时的测试中未发现通用越狱。定价为每百万输入令牌 10 美元、输出令牌 50 美元，有限计划在 6 月 22 日前可免费使用。
Key points
Claude Fable 5 is Anthropic's most capable publicly available model, with general availability.
Claude Fable 5 是 Anthropic 最强大的公开可用模型，已全面可用。
Fable 5 migrated a 50-million-line Ruby codebase in a day, saving months of engineering work.
Fable 5 在一天内迁移了 5000 万行 Ruby 代码库，节省了数月的工程时间。
Mythos 5 autonomously conducted genomics research across 138 species, outperforming a Science paper with a model 100x smaller.
Mythos 5 自主进行了跨 138 个物种的基因组研究，以缩小 100 倍的模型超越了《科学》论文。
Safety classifiers fall back to Opus 4.8 on ~5% of sessions, with zero universal jailbreaks found after 1,000+ hours of testing.
安全分类器在约 5% 的会话中回退到 Opus 4.8，经过 1000 多小时测试未发现通用越狱。
Pricing: $10/M input tokens and $50/M output tokens, with free access through June 22 for Pro/Max/Team plans.
定价：每百万输入令牌 10 美元、输出令牌 50 美元，Pro/Max/Team 计划在 6 月 22 日前可免费使用。