Kimi
FreemiumAI assistant by Moonshot AI, built on the Kimi K3 model with a 1M token context window and native visual understanding
What is Kimi?
Kimi is the consumer AI assistant from Moonshot AI, one of China's leading AI labs. You can use it on the web at kimi.com, through mobile apps on iOS and Android, via the Moonshot / Kimi API for developers, and through the Kimi Code CLI for terminal-based coding workflows. Kimi shipped as a chatbot with bilingual Chinese-English support and quickly became known for long-context document handling and aggressive pricing.
Under the hood, Kimi is powered by the Kimi K3 model, which Moonshot calls its most capable model to date. K3 has 2.8 trillion parameters and native visual understanding, and the vendor serves it through kimi.com, the Kimi App, the Kimi Work desktop app, Kimi Code, and the Kimi API. It ships with a 1M-token context window, long enough to load an entire book, a large codebase, or a stack of research papers into one conversation, and Moonshot also publishes the code-tuned kimi-k2.7-code models and the general-purpose kimi-k2.6, all with 256K-token context windows, plus an "Agent Swarm" mode that coordinates multiple specialized sub-agents to tackle complex tasks in parallel.
What sets Kimi apart in 2026 is the combination of a very long context window and a tiered model lineup. API pricing for Kimi K3 is $3.00 per million input tokens and $15.00 per million output tokens on pay as you go, while the cheapest models in the current lineup are $0.95 per million input tokens and $4.00 per million output tokens. Combined with automatic context caching (which can reduce input costs by up to 75% on repeated prompts), Kimi remains a practical option for long-context work. The consumer Kimi Chat at kimi.com remains free, with paid tiers unlocking higher usage for features like Deep Research and the OK Computer agent.
Kimi demo video
Watch Kimi AI's official demo to see Kimi in action before reading our full review.
Official video by Kimi AI via YouTube, embedded for reference. ToolChase does not host or claim this video.
⚡ Quick Verdict
Researchers and professionals who need to analyze extremely long documents, books, or datasets in a single conversation
Users who prioritize advanced reasoning or need deep English-language ecosystem integrations
Free with daily limits. Premium plans for higher usage.
Yes
1M-token context window on the current Kimi K3 flagship
General reasoning trails frontier Western models like GPT-4o and Claude
Bottom line: Kimi scores 4.1/5, Researchers and professionals who need to analyze extremely long documents, books, or datasets in a single conversation.
Kimi Pricing
Kimi Chat (web and mobile), Free: The Adagio free plan at kimi.com offers unlimited basic conversations with Kimi K3 and file uploads. Intensive features like Deep Research queries and OK Computer agent tasks are rate-limited on the free tier.
Paid Kimi Chat plans: Moonshot offers higher-tier consumer plans that access more Deep Research runs, more agent executions, and larger file limits. Exact tier pricing varies by region and launch market, see the official Kimi pricing page for current monthly rates.
Kimi K3 API: $3.00 per million input tokens and $15.00 per million output tokens on pay as you go. The cheaper current models, kimi-k2.7-code and kimi-k2.6, are $0.95 per million input tokens and $4.00 per million output tokens with 256K-token context windows. The 1M-token context window on K3 is included at no premium, and Moonshot's automatic context caching can reduce input costs by up to 75% for repeated or overlapping prompts, which is particularly useful for long-document RAG and agent workflows.
Note: Always verify current API and subscription pricing on platform.kimi.ai and kimi.com before committing to a plan. Pricing has shifted several times as Moonshot moved from the K2 series to K2.5 and then to K3.
Key Features
- 1M-token context window on Kimi K3: Large enough to ingest a full novel, a 200+ page PDF stack, or a midsize codebase in a single conversation without splitting into chunks.
- Kimi K3 flagship model: 2.8 trillion parameters with native visual understanding, which Moonshot calls its most capable model to date.
- Code-tuned kimi-k2.7-code models: Separate code-focused models with 256K-token context windows, including a highspeed variant, available alongside the K3 flagship.
- Agent Swarm mode: Coordinates multiple specialized sub-agents in parallel on complex tasks, which Moonshot reports can meaningfully reduce wall-clock time versus sequential approaches.
- Automatic context caching: Repeated or overlapping prompts reuse cached context so input costs drop by up to 75%, particularly valuable for RAG and agent loops.
- Multi-file upload and analysis: Upload PDFs, Word documents, spreadsheets, and code files and ask Kimi to cross-reference and summarize them.
- Web search integration: Kimi can pull current information from the web to ground answers and citations.
- Deep Research mode: A longer-running research workflow where Kimi breaks a query into sub-questions, searches, and drafts a structured report.
- OK Computer agent: Moonshot's agent runtime that executes multi-step tasks on behalf of the user, with higher quotas on paid tiers.
- Kimi Code CLI: Terminal-native coding assistant for developers who prefer a command-line workflow.
- Bilingual Chinese-English support: Kimi is particularly strong in Chinese and offers solid English quality as well.
Best For
Developers who need long context on a budget: Kimi K3 is $3.00 / $15.00 per million input/output tokens on pay as you go, and the cheapest current models are $0.95 / $4.00 per million input/output tokens, with automatic context caching on top. Good pick for high-volume RAG, agent loops, and research pipelines.
Researchers and analysts processing long documents: The 1M-token context window on Kimi K3 comfortably handles entire books, long legal contracts, multi-file research projects, and large codebase reviews in a single conversation.
Bilingual and Chinese-market users: Kimi has first-class Chinese language support and is widely used inside China. If you're building for or working with Chinese-language content, Kimi is one of the strongest options.
Users who want to try agentic AI cheaply: Agent Swarm and OK Computer give you a low-cost way to experiment with multi-agent workflows, long-running tasks, and autonomous research without burning through frontier-model budgets.
Pros & Cons
Pros
- 1M-token context window on Kimi K3 handles entire books and large codebases without chunking
- Lower-cost model tiers priced below the K3 flagship for high-volume work
- Automatic context caching cuts repeat-prompt input cost by up to ~75%
- Kimi Chat is free to use on the web and mobile
- Strong bilingual Chinese-English support
- Agent Swarm and OK Computer offer experimental multi-agent workflows
- Code-tuned kimi-k2.7-code models run alongside the K3 flagship
- Kimi Code CLI makes it easy to integrate into developer workflows
Cons
- General English-language polish still trails GPT-5-class and Claude Sonnet 4.x for nuanced writing
- Operated by a China-based company, data residency and compliance concerns for some enterprises
- Ecosystem and third-party integrations are smaller than ChatGPT or Claude
- Documentation and community support are mostly Chinese-first
- Free tier features like Deep Research and OK Computer have tight daily limits
- Context window is long, but not the largest on the market (Qwen3.6-Plus offers 1M tokens)
- Pricing has shifted multiple times as the K2 line evolved
- Some regulated industries will rule Kimi out purely on jurisdiction
FAQ
What is Kimi and who makes it?
Kimi is a consumer AI assistant and model family from Moonshot AI, one of China's leading AI labs. The consumer-facing chatbot is available at kimi.com and on iOS/Android. The current flagship model, Kimi K3, has 2.8 trillion parameters, native visual understanding, and a 1M-token context window. Kimi also ships an Agent Swarm mode for multi-agent workflows. The earlier Kimi K2.5 and Kimi K2 models, including the K2 Thinking variant, have been discontinued by the vendor. Developers can access it via the Moonshot / Kimi API or the Kimi Code CLI.
Is Kimi free to use?
Yes. The Kimi Chat Adagio free plan at kimi.com gives you unlimited basic conversations with Kimi K3 and file uploads. Intensive features like Deep Research queries and OK Computer agent tasks are rate-limited on the free tier. Moonshot offers paid consumer tiers for heavier usage, and a pay-as-you-go API for developers (Kimi K3 is $3.00 per million input tokens and $15.00 per million output tokens, and the cheapest current models are $0.95 per million input tokens and $4.00 per million output tokens), with automatic context caching that can cut repeat-prompt costs by up to 75%.
How big is Kimi's context window really?
The current Kimi K3 flagship supports a 1M-token context window, and the vendor serves K3 through the Kimi App, kimi.com, the Kimi Work desktop app, Kimi Code, and the Kimi API. The older 256K figure belongs to Kimi K2.5, which the vendor has discontinued; 256K is also the context window on the current kimi-k2.7-code and kimi-k2.6 models. A 1M-token window is comfortably larger than ChatGPT's standard 128K and Claude's 200K, and it is enough for book-length documents, multi-file PDF stacks, and large codebases. Gemini and Qwen3.6-Plus also offer 1M-token windows.
How does Kimi compare to ChatGPT, Claude, and DeepSeek?
ChatGPT and Claude still lead on polish, English nuance, and ecosystem, integrations, plugins, enterprise features, and third-party tooling. Kimi wins on price, long-context handling at scale, and strong Chinese-language support. DeepSeek is the other low-cost Chinese option with strong reasoning. For Western English-language SaaS, Claude or ChatGPT remain easier defaults; for cost-sensitive long-document workloads or bilingual Chinese-English tasks, Kimi is very competitive.
Is Kimi good at coding?
Yes, coding is a focus of the current lineup. The 1M-token context window on Kimi K3 means you can load substantial portions of a codebase and ask Kimi for cross-file refactors, code reviews, and bug hunts in a single pass, and Moonshot also publishes the code-tuned kimi-k2.7-code and kimi-k2.7-code-highspeed models with 256K-token context windows. The Kimi Code CLI makes it easy to pipe files and directories into the model directly from your terminal, which makes it attractive for high-volume code-review or agent workflows.
Is Kimi safe for sensitive or regulated data?
Kimi is operated by Moonshot AI, a China-headquartered company, and data you send to its hosted services may be processed in jurisdictions and under regulations that differ from US or EU norms. For confidential, legally privileged, or regulated personal data, and for many enterprise security reviews, that jurisdictional question alone is a blocker, and you should use a US or EU-based alternative. For general-purpose, non-sensitive workloads and public-data research, Kimi is fine, but always read the latest privacy policy and terms before deploying it in production.
What is Kimi Agent Swarm?
Agent Swarm is Kimi's multi-agent execution mode, where a primary model coordinates several specialized sub-agents running in parallel on different parts of a task. Moonshot reports that this parallelism can meaningfully cut wall-clock time on complex jobs compared to a single sequential model, particularly for benchmarks and agentic research tasks. It's one of the more interesting production examples of the "swarm of agents" pattern, and combined with Kimi's low per-token pricing, it gives developers a way to experiment with multi-agent workflows without quickly burning through a frontier-model budget.
Score based on product quality, usability, value, features, reliability, integrations & market trust. How we score →
📋 Good to know
Visit www.kimi.com or download mobile apps. Account creation required. Available globally but optimized for Chinese and English content.
Operated by Moonshot AI (China-based). Data processing follows Chinese regulations. For sensitive documents, consider privacy implications of jurisdiction.
Upgrade when you hit daily free tier limits and need consistent high-volume document processing.
Very low. Standard chatbot interface. The main advantage is simply uploading larger files than competitors allow.