Updated March 25, 2026· Based on independent benchmark data
OpenAI: GPT-5.2-Codex leads in intelligence with a score of 54.0 vs 51.7. OpenAI: GPT-5.2-Codex is 1.7x cheaper at $1.75/1M tokens vs $3.00/1M.
| Metric | OpenAI: GPT-5.2-Codex | Anthropic: Claude Sonnet 4 |
|---|---|---|
| Intelligence Score | 54.0 | 51.7 |
| Coding Score | 53.1 | 50.9 |
| Math Score | N/A | N/A |
| Speed (tok/s) | 68 tok/s | 65 tok/s |
| Latency (TTFT) | 75.14s | 38.68s |
| Input Price / 1M tokens | $1.75 | $3.00 |
| Output Price / 1M tokens | $14 | $15 |
| Context Window | 400K |
OpenAI: GPT-5.2-Codex outperforms Anthropic: Claude Sonnet 4 on the Artificial Analysis intelligence index with a score of 54.0 compared to 51.7. For coding tasks, OpenAI: GPT-5.2-Codex has the edge with a coding score of 53.1 vs 50.9.
Both models deliver similar output speeds: OpenAI: GPT-5.2-Codex at 68 tok/s and Anthropic: Claude Sonnet 4 at 65 tok/s. Time to first token is 38.68s for Anthropic: Claude Sonnet 4 vs 75.14s for OpenAI: GPT-5.2-Codex, which affects perceived responsiveness in interactive applications.
OpenAI: GPT-5.2-Codex is more affordable at $1.75/1M input tokens ($14/1M output), while Anthropic: Claude Sonnet 4 costs $3.00/1M input ($15/1M output). For a typical workload of 100 requests per day at 2,000 tokens each, OpenAI: GPT-5.2-Codex would cost approximately $10.50/month vs $18.00/month for Anthropic: Claude Sonnet 4 in input costs alone.
OpenAI: GPT-5.2-Codex offers a larger context window at 400K tokens compared to Anthropic: Claude Sonnet 4's 200K. For output length, OpenAI: GPT-5.2-Codex can generate up to 128K tokens per response vs 64K for Anthropic: Claude Sonnet 4.
Choose OpenAI: GPT-5.2-Codex when you need higher intelligence (54.0), stronger coding performance (53.1), lower cost.
OpenAI: GPT-5.2-Codex scores higher on coding benchmarks (53.1 vs 50.9), making it the better choice for programming tasks.
OpenAI: GPT-5.2-Codex is cheaper at $1.75/1M input tokens vs $3.00/1M for Anthropic: Claude Sonnet 4.
OpenAI: GPT-5.2-Codex is faster, producing output at 68 tok/s compared to Anthropic: Claude Sonnet 4's 65 tok/s.
Yes, OpenAI: GPT-5.2-Codex supports image input. Anthropic: Claude Sonnet 4 also supports images.
Benchmark data by Artificial Analysis
Data last synced: March 25, 2026
| 200K |
| Max Output Tokens | 128K | 64K |
| Input Modalities | Text + Image | Image + Text + File |
| Output Modalities | Text | Text |
| Free Tier | No | No |
OpenAI: GPT-5.2-Codex has a larger context window at 400K compared to Anthropic: Claude Sonnet 4's 200K.
It depends on your priorities. OpenAI: GPT-5.2-Codex scores higher on intelligence (54.0), but Anthropic: Claude Sonnet 4 may be better for specific use cases like budget-conscious projects or speed-critical applications.