Comparison
GPT-5.3-Codex vs Claude Opus 4.8 — cost compared
GPT-5.3-Codex and Claude Opus 4.8, compared on real per-million-token pricing.
| Attribute | GPT-5.3-Codex | Claude Opus 4.8 |
|---|---|---|
| Input / 1M | $1.75 | $5.00 |
| Output / 1M | $14.00 | $25.00 |
| Cache read / 1M | $0.18 | $0.50 |
| Cache write 1h / 1M | — | $10.00 |
Reading the comparison
All figures are USD per million tokens at API list prices. Cache-read economics often matter more than the headline input/output rates for a well-cached coding session.
Common questions
- Which is cheaper, GPT-5.3-Codex or Claude Opus 4.8?
- It depends on your input/output mix and how much you cache. The table shows each rate so you can weigh your own workload.