Kimi K2.5 Thinking API

Model ID: kimi-k2.5-thinking · Moonshot · Billed per token

Kimi K2.5 Thinking pricing

AibermOfficial
Input$0.414 / 1M tokens$0.600
Output$2.07 / 1M tokens$3.00
Cache read$0.083 / 1M tokens

Prices in USD. About 31% below the official price. Pay as you go, no minimum spend.

Get an API key · Read the docs

How to call Kimi K2.5 Thinking

Use your Aiberm API key, set the base URL, and pass kimi-k2.5-thinking as the model name.

Supported request formats:

curl https://aiberm.org/v1/chat/completions \
  -H "Authorization: Bearer $AIBERM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "kimi-k2.5-thinking", "messages": [{"role": "user", "content": "Hello"}]}'

Use Kimi K2.5 Thinking in your tools

FAQ

How much does the Kimi K2.5 Thinking API cost?

On Aiberm, Kimi K2.5 Thinking (kimi-k2.5-thinking) costs $0.414 per 1M input tokens and $2.07 per 1M output tokens, compared with the official $0.600 / $3.00 — about 31% less. You pay only for the tokens you use, with no minimum spend.

How do I call Kimi K2.5 Thinking through Aiberm?

Set the base URL to https://aiberm.org, use kimi-k2.5-thinking as the model name, and authenticate with your Aiberm API key. Supported request formats: Anthropic Messages, OpenAI Chat Completions.

Does Kimi K2.5 Thinking support prompt caching?

Yes. Cached input tokens are billed at $0.083 per 1M tokens.

Can I use Kimi K2.5 Thinking in Claude Code, OpenCode or OpenClaw?

Yes. Add Aiberm as a provider in Claude Code, OpenCode, OpenClaw with the base URL and your API key, then select kimi-k2.5-thinking. Each tool has a step-by-step guide in the Aiberm docs.

More Moonshot models

All models and prices