signalAI热榜2026-09-24
OpenRouter Analyzes Kimi K3: Open Weights, License Terms, and How to Call It
Kimi K3 is open-weights, not open source, under Moonshot AI's custom license that adds conditions for large-scale use: a $20M revenue threshold for Model-as-a-Service and a UI attribution requirement for products with over 100M MAU. The checkpoint is a 2.8-trillion-parameter Mixture of Experts model with 104 billion active parameters per token, stored in MXFP4. On OpenRouter, model ID moonshotai/kimi-k3 supports text, image, video inputs, a 1,048,576-token context, reasoning effort, tools, and structured outputs, with pricing from $1.80 to $6.00 per million prompt tokens.
- for who
- Developers who want to call Kimi K3 through OpenRouter or self-host it.
- why now
- Kimi K3's open-weight release and API pricing are newly available, so act before licensing terms change.
- what changes
- They must comply with license terms for large-scale use, and can access the model via OpenRouter with flexible reasoning controls.
- to do
- Use model ID moonshotai/kimi-k3 on OpenRouter, set reasoning.effort and consider provider.require_parameters for endpoint filtering.
key points
- Kimi K3 is open-weights, not open source; custom license with revenue and attribution thresholds
- 2.8T parameter MoE, 104B active per token, MXFP4, context 1,048,576 tokens
- OpenRouter model moonshotai/kimi-k3, pricing varies by endpoint, supports tools and reasoning
#Kimi K3#open weights#license#api call
score
score 8 out of 10. 0-10: how dense the facts are, multiplied by how much you can do with them after reading. 8+ means the topic's evidence bar is met: benchmarks and availability for a new model, amount and investors for a funding round, revenue figures for a solo-money story. Below 5 an item does not enter the digest. A press release scores 3 or less, a reprint loses 2, anything older than 14 days loses 1, a headline that misleads loses 3.
read the source