Moonshot Kimi K3.1 Pro
Moonshot's flagship open-weight multimodal reasoning model, post-trained from Kimi K3 for long-horizon coding and agent work, with a 1M-token context.
Kimi K3.1 Pro is a cloud text model built by Moonshot. It is multimodal: alongside a text prompt it accepts images, then replies with generated text. Its context window handles up to 1.05M input tokens and up to 131K output tokens, and it supports adjustable reasoning effort for harder problems. Generation can be tuned with system instructions. It runs through OpenRouter using your own API key, from $2.00 per million input tokens.
- Released
- Oct 2026
- Pricing
- $2.00 / 1M in · $10.00 / 1M out
- Inputs
- Text, Images
- Context window
- 1.05M in · 131K out
- Reasoning
- Adjustable effort
- Controls
- System prompt
Samples
Generated with Kimi K3.1 Pro via OpenRouter. The same three prompts run against every text model in the catalog, shown verbatim, so the only thing that changes between two models’ answers is the model.
Explain what an API is to someone who has never written code. Do not use a restaurant or waiter analogy. Plain prose only — no headings, bullet points, or markdown formatting. Maximum 50 words.
An API, or application programming interface, is a set of rules that lets one program request information or services from another. Like a power outlet, it offers a standard connection: you plug in correctly and get what you need without knowing how the system behind it works.
A two-person startup is deciding whether to build their own user authentication or use a hosted provider. Give them a clear recommendation, then the single strongest argument against your own recommendation. Plain prose only — no headings, bullet points, or markdown formatting. Maximum 150 words.
Use a hosted authentication provider. With only two people, your scarce time is better spent on the product that differentiates you, not on password storage, session handling, account recovery, multifactor authentication, and abuse defenses. A reputable provider gives you a safer baseline, a faster launch, and fewer opportunities for a catastrophic security mistake. You can revisit the decision if your scale or requirements change. The strongest argument against this is vendor lock-in: authentication touches every user and many core flows, so migrating identities, credentials, and integrations later can be costly and disruptive. If the provider’s pricing or terms become unfavorable, you may have little leverage precisely when your business depends on it.
Write a briefing for a non-technical executive explaining why their company's spending on AI is hard to predict, and what they can do about it. Cover at least three distinct causes. Plain prose only — no headings, bullet points, or markdown formatting. Write exactly 205 words.
Your AI bill is hard to forecast because it behaves less like a software subscription and more like electricity. Many services charge for each unit of text processed, so costs rise with customer demand, document length, and the number of attempts needed to get a useful answer. A small change in a product feature or employee habit can multiply usage. Second, the price of a model is only part of the expense. Data preparation, security reviews, testing, human oversight, and integration work often arrive separately and vary by use case. Third, the market keeps moving: vendors change prices and limits, while newer models may alter both capability and consumption. Success creates uncertainty, since a cheap pilot can become expensive when widely adopted. Treat AI as a portfolio of business investments, not a technology budget. For each use case, measure cost per task alongside quality and revenue or time saved. Start with capped pilots, require ownership, and tag spending by product and team. Set budgets, usage alerts, and approval thresholds before scaling. Compare models, use cheaper options for routine work, and avoid commitments until demand is stable. Review forecasts monthly using low, expected, and high usage scenarios, and stop projects whose unit economics do not improve.
Moonshot
Moonshot builds the Kimi family — frontier open-weight models with very long context, native visual understanding, and strong agentic coding.
www.moonshot.ai ↗Moonshot's 2.8T-parameter open-weight mixture-of-experts reasoning model (104B active) with a 1M-token context, built for long-horizon agent work, coding and research.
Moonshot's frontier open-weight model (1T parameters) built for long-horizon agentic coding, with a 262K context and native vision.