moonshotai Open weights Reasoning

Kimi K2 Thinking pricing & benchmarks

Released Nov 6, 2025 · served by 2 providers

Input / 1M tokens$0.600
Output / 1M tokens$2.50
Context window262K
Cached input$0.150
Intelligence index17.3
Coding index21.0
Agentic index1.8
1M in + 300K out$1.35

What Kimi K2 Thinking is

Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...

Providers and prices for Kimi K2 Thinking

The same model costs different amounts depending on who serves it. Prices are per 1M tokens, cheapest first. Uptime and throughput are OpenRouter's rolling measurements, not vendor claims.

ProviderInputOutputContext QuantUptime 24hThroughput
Novitacheapest $0.600 $2.50 262K bf16 99.5%
Google $0.600 $2.50 262K 95.8%

Kimi K2 Thinking benchmark results

Design Arena head-to-head results, as reported through the OpenRouter model API.

CategoryArenaRankEloWin rate
website models #79 1138 48.8%

Cheaper models in the same class

Models scoring within 4 points of Kimi K2 Thinking on the Intelligence Index, but with a lower output price.

FAQ

How much does the Kimi K2 Thinking API cost?

$0.600 per 1M input tokens and $2.50 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $1.35.

Is Kimi K2 Thinking free?

No. It is a paid model starting at $0.600 per 1M input tokens, though some providers offer trial credits.

Which provider is cheapest for Kimi K2 Thinking?

Novita at $0.600 input / $2.50 output per 1M tokens (bf16 quantization).

Can I self-host Kimi K2 Thinking?

Yes — weights are published as moonshotai/Kimi-K2-Thinking on Hugging Face, so you can run it on your own hardware.