Vercel

Anthropic's Claude Sonnet 4.6 Launches on AI Gateway with Agentic Capabilities


Executive Summary

Anthropic has released its new AI model, Claude Sonnet 4.6, which is now available on the AI Gateway platform. This model features a 1 million token context window and approaches the intelligence of Anthropic's top-tier Opus model, with significant improvements in agentic coding, code review, and user interface generation. Designed for developers, Sonnet 4.6 is a hybrid reasoning model that can deliver both near-instant responses and perform extended thinking within the same API call.

Key Takeaways

* Model: Claude Sonnet 4.6 from Anthropic.

* Availability: Immediately available on AI Gateway.

* Performance: Approaches "Opus-level intelligence" with a 1 million token context window.

* Key Capabilities: Shows strong improvements in agentic coding, code review, frontend UI quality, and computer use accuracy.

* Agentic Features: Can proactively execute tasks, delegate to subagents, and parallelize tool calls.

* Hybrid Reasoning: Combines near-instant responses with extended thinking in a single model, controllable via API parameters like `effort` and `thinking: { type: 'adaptive' }`.

* Implementation: Developers can access the model via the AI SDK by specifying `anthropic/claude-sonnet-4.6`.

Strategic Importance

This launch provides developers using AI Gateway with a more powerful and versatile model that bridges the gap between speed and top-tier intelligence, enabling the creation of more sophisticated and autonomous AI agents.

Original article