Head to head
Claude Haiku 4.5 vs GPT-5.6 Luna
On a blended workload of one million input tokens and three million output tokens, GPT-5.6 Luna is 4.2× cheaper than Claude Haiku 4.5.
| Claude Haiku 4.5 | GPT-5.6 Luna | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Input / MTok | $1 | $0.2 |
| Output / MTok | $5 | $1.2 |
| Cache read / MTok | $0.1 | $0.02 |
| Context window | 200,000 | — |
| Blended 1M in + 3M out | $16.00 | $3.80 |
| Status | ga | ga |