Model Comparator

Compare up to four models using the same criteria.

GPT-5 Codex (High)
OpenAI
Claude Sonnet 4.6 (Non-reasoning, High)
Anthropic

Differences only

✓ Best published value among measured models. — Data unavailable. These metrics do not establish an overall ranking.

Model Comparator
Metrics	GPT-5 Codex (High) OpenAI	Claude Sonnet 4.6 (Non-reasoning, High) Anthropic
Intelligence	Best: 24.9	24.7
Coding	—	63
Math	98.7	—
Agentic	—	31.8
Speed	Best: 166 t/s	45.2 t/s
TTFT	6.71 s	Best: 1.08 s
Input · $ / 1M tokens	Best: $1.25	$3×2.4 relative to the lowest price among the compared models
Output · $ / 1M tokens	Best: $10	$15×1.5 relative to the lowest price among the compared models
Context window	400K	Best: 1M
Reasoning	Yes	No
Open weights	No	No

Intelligence

GPT-5 Codex (High): Best: 24.9
Claude Sonnet 4.6 (Non-reasoning, High): 24.7

Coding

GPT-5 Codex (High): —
Claude Sonnet 4.6 (Non-reasoning, High): 63

Math

GPT-5 Codex (High): 98.7
Claude Sonnet 4.6 (Non-reasoning, High): —

Agentic

GPT-5 Codex (High): —
Claude Sonnet 4.6 (Non-reasoning, High): 31.8

Speed

GPT-5 Codex (High): Best: 166 t/s
Claude Sonnet 4.6 (Non-reasoning, High): 45.2 t/s

TTFT

GPT-5 Codex (High): 6.71 s
Claude Sonnet 4.6 (Non-reasoning, High): Best: 1.08 s

Input · $ / 1M tokens

GPT-5 Codex (High): Best: $1.25
Claude Sonnet 4.6 (Non-reasoning, High): $3×2.4 relative to the lowest price among the compared models

Output · $ / 1M tokens

GPT-5 Codex (High): Best: $10
Claude Sonnet 4.6 (Non-reasoning, High): $15×1.5 relative to the lowest price among the compared models

Context window

GPT-5 Codex (High): 400K
Claude Sonnet 4.6 (Non-reasoning, High): Best: 1M

Reasoning

GPT-5 Codex (High): Yes
Claude Sonnet 4.6 (Non-reasoning, High): No

Open weights

GPT-5 Codex (High): No
Claude Sonnet 4.6 (Non-reasoning, High): No

Each reasoning level remains a separate configuration.

Score and cost

Each line joins the reasoning levels of a compared model; the large point marks the configuration in the table. The top-right corner holds the best score for the lowest cost. Select another point to compare that level.

GPT-5 Codex (High)24.9 · $3.44
Claude Sonnet 4.6 (Non-reasoning, High)24.7 · $6