The latest iteration of Anthropic's flagship AI model is reportedly heading to the multi-model routing platform, though official confirmation remains elusive.
Anthropic’s Claude Sonnet 5, internally codenamed “Fennec,” is reportedly being prepared for release on OpenRouter, the unified AI inference platform that has become a go-to hub for developers who want access to multiple large language models through a single API.
The timing is notable. Rumors about Sonnet 5 have been circulating since early February 2026, when speculation pointed to a launch around February 3. That didn’t happen. Now, nearly five months later, the model appears to finally be approaching availability, at least through one major distribution channel.
What we know about the rollout #
OpenRouter already hosts several Anthropic models, including Claude Sonnet 4.6, which launched on February 17, 2026. That model is priced at $3 per million input tokens and $15 per million output tokens.
For Sonnet 5, the specifics around pricing, context window size, and performance benchmarks haven’t been publicly disclosed. Neither Anthropic nor OpenRouter has issued a formal announcement confirming availability or release timing. That said, OpenRouter has consistently been among the first third-party platforms to offer new Anthropic models, and its infrastructure has been built to support features like reasoning traces, which allow developers to inspect the step-by-step logic a model uses to arrive at its output.
The long road from rumor to release #
Claude Sonnet 5 has had one of the longer gestation periods in recent AI model history. The initial February 2026 rumors suggested that internal issues delayed the planned rollout, though neither Anthropic nor any official source ever confirmed what those issues were.
The gap between Sonnet 4.6 and Sonnet 5 is roughly four and a half months, assuming the model does drop today. OpenRouter’s multi-model marketplace makes the competitive dynamic across providers especially visible, since developers can directly compare performance and pricing across providers in real time.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our