Kotak Institutional Equities reported that Anthropic's release of the Claude Fable 5 model increases AI disruption risks for IT services firms. The new model, a better version of the Mythos Preview model announced in April, shows a significant step up in software engineering tasks, increasing revenue deflation risks for IT services firms in the near to medium term, particularly for those with large exposure to application services. Claude Fable 5 is publicly available and similar to Mythos 5, but has guardrails to restrict usage in certain areas; Mythos 5 lacks these guardrails and is available only to select partners. The models have a one million context window, with pricing at 2X of Opus 4.8 and higher than other frontier models, but ~2-2.5X lower than Mythos Preview. In SWE Bench Pro, Fable 5 scores 11% higher than Opus 4.8 and 22% higher than GPT 5.5. In Frontier Code (Diamond), it scores 16% higher than Opus 4.8 and 24% higher than GPT 5.5. METR’s analysis on the Claude Mythos Preview model estimated it could complete 3 hour+ tasks with an 80% success rate. The model does not seem to exhibit a similar capability improvement in other domains. Factors constraining widespread enterprise adoption include a focus on controlling token costs.