gpt-5.6-terra vs gpt-5.6-luna
gpt-5.6-terra vs gpt-5.6-luna: $2/$12 vs $0.2/$1.2 per 1M, the same workload priced on both; gpt-5.6-luna is 90% cheaper per request.
| gpt-5.6-terra | gpt-5.6-luna | |
|---|---|---|
| Provider · tier | openai · mid | openai · fast |
| Input | $2 | $0.2 |
| Cached input | $0.2 | $0.02 |
| Cache write | $2.5 | $0.25 |
| Output | $12 | $1.2 |
| Long-context tier | > 272,000: $4 / $18 | > 272,000: $0.4 / $1.8 |
| Per request (2,000 in / 300 out) | $0.0076 | $0.00076 |
| Per month at 100k requests | $760 | $76 |
| …with the static prefix cached | $490 | $49 |
| Cheaper sibling (cascade) | gpt-5.6-luna | — |
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, gpt-5.6-terra or gpt-5.6-luna?
At list prices (table 2026-10-26) gpt-5.6-luna is 90% 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.