Head to head
Claude Haiku 4.5 vs GPT-5.6 Luna
Cost per solved task needs a published pass rate for both models. At least one of these has none, so it is reported as missing rather than estimated.
| Claude Haiku 4.5 | GPT-5.6 Luna | |
|---|---|---|
| Input $/Mtok | $1 | $0.2 |
| Output $/Mtok | $5 | $1.2 |
| Cached in $/Mtok | $0.1 | $0.02 |
| Pass rate | missing | missing |
| Cost per solved task | missing | missing |
Heavy task tier. Prices verified 2026-08-21.Methodology ·Claude Haiku 4.5 ·GPT-5.6 Luna