jishie

Kimi K2 Thinking

qiniu-ai/moonshotai/kimi-k2-thinking · in: text → out: text · price provider-declared (via models.dev). Capabilities: tools text.

This is a text model — providers price it per unit, not per token, so models.dev carries no token price (fields read “—”). See Qiniu’s docs for per-unit rates.

Provider
Qiniu (qiniu-ai)
Kind
text (outputs text)
Context
256,000 tokens
Max output
100,000 tokens
Capabilities
tools text

Same model, other providers

“Kimi K2 Thinking” is offered by 16 providers — cheapest-effective first. Vercel AI Gateway is the cheapest at $0.85 blended.

ProviderInputOutputBlendedContext
Vercel AI Gateway$0.47$2$0.85216k
Helicone$0.48$2$0.86256k
OpenCode Zen$0.4$2.5$0.93262k
IO.NET$0.55$2.25$0.9833k
NanoGPT$0.6$2.5$1.07262k
Kilo Gateway$0.6$2.5$1.07262k
NovitaAI$0.6$2.5$1.07262k
OpenRouter$0.6$2.5$1.07262k
ZenMux$0.6$2.5$1.07262k
Charm Hyper$0.6$2.5$1.07262k
Merge Gateway$0.6$2.5$1.07262k
Amazon Bedrock$0.6$2.5$1.07262k
Qiniu (this)———256k

Call it

jishie indexes & prices models; it does not proxy inference.

Most providers are OpenAI-compatible — point base_url at Qiniu and pass this model id:

curl $BASE_URL/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" -H "Content-Type: application/json" \
  -d '{"model":"moonshotai/kimi-k2-thinking","messages":[{"role":"user","content":"hello"}]}'

Route programmatically with MCP find_model / get_model, or fetch this record free (no wallet):

← all Qiniu models · model index · JSON