OpenAI GPT-5 ($1.25 / $10) vs Anthropic Claude Opus 4.7 ($5 / $25) per 1M tokens.
GPT-5 and Claude Opus 4.7 sit at very different points on the price-capability curve. GPT-5 is OpenAI's reliable production workhorse at $1.25 / $10 per million tokens. Claude Opus 4.7 is Anthropic's flagship reasoning model at $5 / $25 per million tokens — 4× more on input, 2.5× more on output.
On most chat and coding workloads, GPT-5 ships answers that are within a few percentage points of Opus 4.7 quality for a fraction of the cost. For long multi-step reasoning, Opus 4.7's quality premium is real and often worth the spread.
GPT-5 is significantly cheaper: $1.25 / $10 vs Opus 4.7 at $5 / $25 per million tokens. For a typical 4:1 input/output workload, GPT-5 is roughly 3.3× less expensive.
Opus 4.7 typically scores higher on agentic coding benchmarks like Terminal-Bench. GPT-5 wins on price-per-correct-answer for most everyday coding tasks. Most teams route easy coding turns to GPT-5 and escalate to Opus 4.7 only for complex refactors.
Default to GPT-5 unless you have a clear quality gap to close. If you measure a meaningful uplift on Opus 4.7, route only your hardest 5–10% of turns to it.