grok-4.7 vs deepseek-v4-pro
grok-4.7 vs deepseek-v4-pro: $2/$6 vs $1.32/$3.96 per 1M, the same workload priced on both; deepseek-v4-pro is 34% cheaper per request.
| grok-4.7 | deepseek-v4-pro | |
|---|---|---|
| Provider · tier | xai · frontier | deepseek · mid |
| Input | $2 | $1.32 |
| Cached input | $0.5 | $0.044 |
| Cache write | none | none |
| Output | $6 | $3.96 |
| Long-context tier | > 200,000: $4 / $12 | none |
| Per request (2,000 in / 300 out) | $0.0058 | $0.003828 |
| Per month at 100k requests | $580 | $383 |
| …with the static prefix cached | $355 | $191 |
| Cheaper sibling (cascade) | grok-4.3 | deepseek-flash |
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, grok-4.7 or deepseek-v4-pro?
At list prices (table 2026-10-26) deepseek-v4-pro is 34% 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.