Head to head
Claude Opus 4.8 vs GPT-5.6 Sol
On a blended workload of one million input tokens and three million output tokens, Claude Opus 4.8 is 1.2× cheaper than GPT-5.6 Sol.
| Claude Opus 4.8 | GPT-5.6 Sol | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Input / MTok | $5 | $5 |
| Output / MTok | $25 | $30 |
| Cache read / MTok | $0.5 | $0.5 |
| Context window | 1,000,000 | — |
| Blended 1M in + 3M out | $80.00 | $95.00 |
| Status | ga | ga |