AI Benchmark Model

Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

Benchmark scores, pricing, speed, and model comparisons for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort).

Key scores

Intelligence51.7Index score
Coding50.9Index score
Speed66.6Tokens/sec
Blended Cost$6.56Per 1M tokens

Claude Sonnet 4.6, developed by Anthropic and released on February 17, 2026, is an AI model featuring Adaptive Reasoning and Max Effort capabilities. The model achieves an intelligence index of 51.7 and a coding index of 50.9. Performance metrics indicate an output speed of 67.675 tokens per second with a time to first token of 53.09 seconds. Pricing is set at $3.75 per million input tokens and $15.00 per million output tokens, resulting in a blended rate of $6.56 per million tokens. Benchmark results include 0.875 on GPQA, 0.3 on HLE, 0.468 on SciCode, 0.565986394557823 on IFBench, 0.706666666666667 on LCR, 0.53030303030303 on TerminalBench Hard, and 0.757309941520468 on TAU2.

Benchmark results

7 tracked benchmark rows for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort).

GPQA Reasoning
87.5%
TAU2 Other
75.7%
LCR Other
70.7%
IFBench Other
56.6%
TerminalBench Hard Other
53.0%
SciCode Other
46.8%
HLE Other
30.0%

Comments (0)

No comments yet

Be the first to share your thoughts!