Claude 3 Haiku vs GPT-5.1-Codex
Anthropic vs OpenAI — benchmarks, pricing, and capabilities side by side.
- •GPT-5.1-Codex has the higher intelligence index (88.2 vs 38.9)
- •Claude 3 Haiku is cheaper ($0.25 vs $1.25 per 1M input)
- •GPT-5.1-Codex is faster
- •GPT-5.1-Codex has a larger context window (400K)
| Claude 3 Haiku | GPT-5.1-Codex | |
|---|---|---|
| Intelligence index | 38.9 | 88.2 |
| Developer | Anthropic | OpenAI |
| Type | Multimodal | Multimodal |
| Access | API only | API only |
| Context window | 200,000 tokens | 400,000 tokens |
| Input price | $0.25 / 1M | $1.25 / 1M |
| Output price | $1.25 / 1M | $10.00 / 1M |
| Speed | 104 tok/s | 188 tok/s |
| Released | March 13, 2024 | November 13, 2025 |
| Parameters | — | — |
| Input modalities | Text, Image | Text, Image |
| Output modalities | Text | Text |
Shared benchmarks
Claude 3 Haiku
GPT-5.1-Codex
GPQA Diamond
33.3
86
Humanity’s Last Exam
3.9
23.4
LiveCodeBench
15.4
84.9