Claude Fable 5.2 lands on AI Gateway: it runs long tasks unattended and hands simpler work to smaller models
Synopsis
Anthropic's Claude Fable 5.2 is now callable through AI Gateway, built for open-ended technical work and improving on earlier Fable models in coding, reasoning and vision, and able to run long, ambiguous tasks unattended. It can act as an orchestrator that hands simpler tasks to smaller models, or as an advisor other Claude models consult on harder problems, with clearer communication and faster output. The model ships with cybersecurity and biology safety classifiers, and access requires data retention: prompts and completions are kept for 30 days and not used to train Claude, while teams with Zero Data Retention enabled get an error instead of a forwarded request.
Interpretation
Claude Fable 5.2 is now available through AI Gateway, positioned for open-ended technical work and improving on earlier Fable models in coding, reasoning and vision. Earlier Fable models are the comparison point, and the new version improves on them in coding, reasoning and vision. The comparison is stated against earlier Fable models across coding, reasoning and vision, not as a head-to-head against other vendors' models.
The model can run long, ambiguous tasks unattended and switch between two roles: dispatching simpler tasks to smaller models, or serving as an advisor that other Claude models consult on harder problems. The orchestrator and advisor roles route work outward rather than having one model carry everything alone. Both roles are described explicitly in the release notes for the open-ended technical work setting.
It is callable through the AI SDK, Chat Completions, Responses and Anthropic Messages APIs, or from a coding agent connected to AI Gateway, under the model name anthropic/claude-fable-5.2. Effort runs from low to max, with high as the API default. The example request falls back to Opus 5.5 and then Sonnet 5.5 when the cybersecurity classifier declines it, and the models option behaves the same on every AI Gateway API format.
Perspective
The release notes cover how to call Claude Fable 5.2 through AI Gateway, the available interfaces, effort levels and fallback configuration, aimed at developers and engineering teams integrating the model; no follow-up experiments or evaluation plans are proposed.
The coding, reasoning and vision improvements are stated relative to earlier Fable models, without specific benchmarks or figures. Access requires data retention; Anthropic does not support Zero Data Retention for this model, prompts and completions are retained for 30 days and not used to train Claude, and teams with Zero Data Retention enabled receive an error instead of having requests forwarded to Fable 5.2.