claude-opus-5-5 vs claude-sonnet-5-5
claude-opus-5-5 vs claude-sonnet-5-5: $4/$20 vs $2/$10 per 1M, the same workload priced on both; claude-sonnet-5-5 is 50% cheaper per request.
| claude-opus-5-5 | claude-sonnet-5-5 | |
|---|---|---|
| Provider · tier | anthropic · frontier | anthropic · mid |
| Input | $4 | $2 |
| Cached input | $0.2 | $0.1 |
| Cache write | $5 | $2.5 |
| Output | $20 | $10 |
| Long-context tier | none | none |
| Per request (2,000 in / 300 out) | $0.014 | $0.007 |
| Per month at 100k requests | $1,400 | $700 |
| …with the static prefix cached | $830 | $415 |
| Cheaper sibling (cascade) | claude-sonnet-5-5 | claude-haiku-5-5 |
Through TokenZero
Point your SDK at the gateway and keep both keys; the cache formatter, response cache and advisory routing act on every request, and each one comes back with a signed receipt showing the list-price baseline, actual cost and the saving.
OPENAI_BASE_URL=https://tokenzero.sh/w/<key>/v1 ANTHROPIC_BASE_URL=https://tokenzero.sh/w/<key>
Questions
Which is cheaper, claude-opus-5-5 or claude-sonnet-5-5?
At list prices (table 2026-10-26) claude-sonnet-5-5 is 50% cheaper per request on a 2,000-in / 300-out workload. Quality is a separate question: TokenZero's cascade only routes to a sibling it scores as sufficient and measures parity with shadow evaluation.
Where do these prices come from?
The providers' own pricing pages, re-verified on 2026-10-26; the gateway signs receipts with the same table.