Home · Technology · Nov 13 archive
Anthropic Launches Claude Fable and Mythos 5.2 Models
Confirmed
In Short: Anthropic has released Claude Fable and Mythos 5.2 models, designed to handle complex technical challenges.
Claude Fable 5.2 is Insane... These are the 9 HUGE Use Cases. Jack Roberts
Anthropic has unveiled Claude Fable and Mythos 5.2, the latest versions of its AI models designed to tackle the most daunting technical challenges, according to the company.
These new models offer improvements in coding, reasoning, and vision, with outputs generated up to 30 percent faster than their predecessors.
Fable 5.2 outperforms Fable 5.1 in agentic coding benchmarks, achieving a score of 49.1% at 'Xhigh effort' compared to 32.4% for Fable 5.1.
In pre-release testing, the models demonstrated their capabilities by designing three working electronic devices from scratch, solving open problems in black hole physics, and formalizing the proof of Fermat’s Last Theorem in 30 hours.
Claude Fable 5.2 and Mythos 5.2 communicate more clearly than their 5.1 versions, with outputs generated up to 30 percent faster and 55 percent fewer blocks.
The cybersecurity features of Mythos 5.2 are currently locked to a specific group: Anthropic’s Cyber Verification Program.
Fable 5.2 is particularly adept at scientific research, scoring 74.6% on Terminal-Bench-Science 0.1, a 22% improvement over Fable 5.1.
Claude Mythos 5.2 identified 10.2% more vulnerabilities than Claude Opus 5.5 in CVE Rediscovery assessments.
The new models are available at a lower price: $8 per million input tokens and $40 per million output tokens, a reduction from the previous $10 per million input tokens.
Anthropic CEO Dario Amodei publicly argued in September that the industry should slow the pace of AI development, and this release marks the company's first since then.
The US government and third-party evaluators reviewed and provided feedback on the new models prior to their launch.
Fable 5.2 and Mythos 5.2 also come with an optional fast mode, which speeds up token generation by up to 2.5 times.
Background
Anthropic announced the launch of Claude Meetings, a new feature that allows its AI assistant Claude to join Google Meet video conferences as a visible participant.
Anthropic has launched Personal Finance in Claude, a feature that allows US subscribers to securely connect and analyze their financial data.
What's confirmed
- Anthropic says Fable 5.2 and Mythos 5.2 communicate more clearly than the 5.1 versions, with outputs generated up to 30 percent faster.
- According to the AI company, Fable 5.2 and Mythos 5.2 were designed to "tackle the most daunting technical challenges" and present "improvements in areas like coding, reasoning, and vision."
- During prerelease testing, the two models have “helped our engineers complete otherwise-infeasible software migrations, designed three working electronic devices from scratch, solved two open problems in the physics of black holes, made substantial progress on robotics evaluations, and formalized the proof of Fermat’s Last Theorem in 30 hours,” Anthropic says.
- In CVE Rediscovery assessments, Claude Mythos 5.2 identified 10.2% more vulnerabilities than Claude Opus 5.5.
- Ad The models produce output up to 30 percent faster than Fable 5.1 and Mythos 5.1, Anthropic says.
- "Fable has strong cybersecurity and biology safeguards to guard against misuse, which allows us to make it available to all users. Organisations pursuing beneficial applications in cybersecurity or biomedical research can apply for access to Mythos through our recently expanded Life Sciences and Cyber Verification Programs," Anthropic said.
- At Medium effort, Fable 5.2 matches Opus 5.5’s top score for about 40% of the cost per task.
- They generate outputs up to 30% faster than Claude Fable 5.1 and Claude Mythos 5.1, and they’re available at a lower price: $8 per million input tokens and $40 per million output tokens.
- AI policy Anthropic has launched Claude Fable 5.2 and Claude Mythos 5.2, two releases built on the same underlying model for open-ended technical work.
What's still developing
- Fable 5.2 outperforms Fable 5.1 in one agentic coding benchmark at 49.1% versus 32.4% at “Xhigh effort,” bringing a meaningful distance in performance.
- Anthropic says that the new version of Fable will return responses 30% faster across effort levels.
- In July, Moonshot AI's Kimi K3 climbed to third place globally on the Artificial Analysis leaderboard, trailing only Claude Fable 5 and GPT-5.6.
- The model is open-weight, priced at roughly a third to a fifth of its US frontier rivals, and reportedly cost a fraction of what OpenAI or Anthropic spend to train comparable systems.
- Claude now writes over 90 percent of the changes merged into Anthropic's core codebase, under human guidance, the company says.
- Last Week in AI #343 - GPT-6, OpenAI’s agents chatted on a wiki, Fable 5.1 Last Week in AI Subscribe Sign in News Last Week in AI #343 - GPT-6, OpenAI’s agents chatted on a wiki, Fable 5.1 GPT-6 Astra Is Here, Discovery of a new OpenAI agent message board, Anthropic launches Claude Fable 5.1 and says it’s up to 45 percent cheaper for agentic work, and more!
- In a new containment test, the model tried to cross boundaries about 85% less often than Opus 5 or Claude Mythos 5.1, with every attempt rated low severity and self-reported.
- It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.
- All of these scores come from Anthropic’s published table, and the company ran the RoboDojo Sim tests internally.
- One Anthropic engineer had Fable 5.2 coordinate over 1,000 subagents for eight days.
- Anthropic has officially launched its newest artificial intelligence models, introducing Claude Fable 5.2 and Claude Mythos 5.2 to the global market.
- Fable 5.2 caters to general users, featuring reinforced safety protocols in biology and cybersecurity while experiencing 55% fewer unnecessary false-positive content blocks than its predecessor.