---
id: "thinkingmachines"
site: "thinkingmachines.ai"
gateways: ["requesty", "vercel"]
models: 3
endpoints: 4
tokens_week: 0
token_share: "0.0%"
rank: 157
cheapest_on: "2 of 2 shared models"
cheaper_than: "100% of offers at the median"
uptime_1d: 100.0000
latency_p50_ms: 1813
tokens_per_second: 165
trains_on_some: true
retains_on_some: true
---
# thinkingmachines

`thinkingmachines/inkling`, `thinkingmachines/inkling-256k`, `thinkingmachines/inkling-small`