Recipe ·
Claude Sonnet 4.6
Claude Sonnet 4.6 is a multimodal large language model developed by Anthropic and released on February 25, 2026. Its documented performance includes an MMLU Pro score of 85.0, an Arena Elo rating of 1470, and a SWE-bench verified score of 79.6, positioning it as a competitive reasoning and coding model. The model was commercially priced at $3 per million input tokens and $15 per million output tok
Ingredient list
Invented by Anthropic · 2022-12 · Velocity 3y
“Anthropic's core alignment methodology is Constitutional AI, used to train Claude models.”
alignmenthighInvented by Google · 2022-01 · Velocity 4y
“Claude models are known for their strong reasoning capabilities, often producing step-by-step reasoning in responses.”
reasoningmediumInvented by Google · 2021-09 · Velocity 4y
“Claude models are instruction-tuned to follow user instructions, a standard practice for Anthropic.”
trainingmediumInvented by Google · 2017-06 · Velocity 9y
“Claude Sonnet is a Transformer-based large language model.”
architecturehigh
This recipe is part of the gentic.news Deployment Atlas. Every ingredient has an origin paper + evidence. Methodology is public. Dataset is CC BY 4.0.