Compare/GPT-5.2 Codex (xhigh) vs Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

GPT-5.2 Codex (xhigh)vsClaude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

Side-by-side comparison of pricing, 12 benchmarks, and generation speed.

OpenAI

GPT-5.2 Codex (xhigh)

Input
$1.75/M
Output
$14/M
Speed
123 tok/s
TTFT
6.69s
Anthropic

Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

Input
$3/M
Output
$15/M
Speed
66 tok/s
TTFT
33.99s

Winner by Category

Cheaper
GPT-5.2 Codex (xhigh)
Faster (tok/s)
GPT-5.2 Codex (xhigh)
Lower Latency
GPT-5.2 Codex (xhigh)
Benchmarks (4-3)
GPT-5.2 Codex (xhigh)

Pricing Comparison

MetricGPT-5.2 Codex (xhigh)Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)
Input ($/M tokens)$1.75$3
Output ($/M tokens)$14$15
Cost for 1M input + 100K output tokens:
GPT-5.2 Codex (xhigh)$3.15
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)$4.50

Speed Comparison

Output Speed (tokens/s) — higher is better
GPT-5.2 Codex (xhigh)
123 tok/s
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)
66 tok/s
Time to First Token (seconds) — lower is better
GPT-5.2 Codex (xhigh)
6.69s
Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)
33.99s

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
49.051.7
Coding Index
43.050.9
Math Index
GPQA Diamond
89.9%87.5%
MMLU-Pro
LiveCodeBench
AIME 2025
MATH-500
Humanity's Last Exam
33.5%30.0%
SciCode
54.6%46.8%
IFBench
77.6%56.6%
TerminalBench
37.1%53.0%
GPT-5.2 Codex (xhigh)4 wins
3 winsClaude Sonnet 4.6 (Adaptive Reasoning, Max Effort)

Frequently Asked Questions

Which is cheaper, GPT-5.2 Codex (xhigh) or Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)?

GPT-5.2 Codex (xhigh) is cheaper overall. Its blended price (3:1 input/output ratio) is $4.81/M tokens vs $6.00/M for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort).

Which model performs better on benchmarks?

GPT-5.2 Codex (xhigh) wins 4 out of 12 benchmarks compared to 3 for Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort). See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

GPT-5.2 Codex (xhigh) generates tokens faster at 123 tok/s vs 66 tok/s. GPT-5.2 Codex (xhigh) also has lower time-to-first-token (6.69s vs 33.99s).

When should I use GPT-5.2 Codex (xhigh) vs Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)?

Choose based on your priorities: GPT-5.2 Codex (xhigh) for lower cost, GPT-5.2 Codex (xhigh) for stronger benchmark performance, and GPT-5.2 Codex (xhigh) for faster generation. For latency-sensitive apps, check the TTFT comparison above.