Kimi Freemium

Kimi

Get bonus
Affiliate link — using it supports this project
+2

Quick facts

Pricing model
Freemium
Minimum price
Free
Free access
Yes
Payment methods
  • Bank card
  • Apple Pay
  • Google Pay
  • Alipay
Service status
Online, checked
User rating
100% positive (2 votes)
Tags
Coding AI agent AI terminal Code generation Bug fixing Code refactoring

Developer and reliability

Information about the developer and the service's terms. This is a factual reference, not a quality assessment or a guarantee of safety.

Developer
Moonshot AI
Stated on the website Source Checked
Domain registered
June 8, 2022
Confirmed by source Source Checked
The domain registration date is not the service's launch date.

Change and verification history

Tracking pricing and information updates

  1. Strengths updated
  2. Limitations updated
  3. Assessment updated
  4. Information verified

Video reviews and tutorials

5

Compare with alternatives

Choose a pair to compare features, pricing and payment methods.

View all alternatives Kimi

About the tool

Standard AI chatbots behave like single-turn responders: you enter a query, the model executes a linear script, and the interaction ends. Kimi is built as an autonomous multi-agent operating environment where complex tasks are divided, delegated, and resolved through multi-threaded execution loops. Developed by Moonshot AI, the workspace integrates deep web research, software engineering, document intelligence, and interactive application deployment.

Instead of forcing users to manually copy and paste code or synthesize documents across disconnected tabs, the platform coordinates specialized agents across a structured toolchain:

  • Autonomous multi-agent orchestration: Through Swarm and Goal modes, the system breaks high-level objectives into modular branches, spawning parallel sub-agents to search web sources, test code snippets, and resolve edge cases until the target milestone is met.
  • Full-stack application and game creation: The Build environment converts natural-language prompts, wireframes, and UI mockups into live web services, interactive 3D simulations, and multi-file codebases complete with database integration and one-click cloud hosting.
  • Deep analytical research: The Deep Research engine crawls hundreds of independent web sources, cross-references findings, and produces long-form structured reports with embedded data tables and analytical charts.
  • Office productivity workspaces: Dedicated modules handle presentation design (Slides) with customizable graphics, spreadsheet calculations (Sheets) with script-based batch processing, and multi-format document conversion (Docs).
  • Developer toolkits and scheduling: Features terminal-ready CLI utilities, IDE plugins for code refactoring, isolated code-execution sandboxes, and automated background schedulers that push reports directly to workspace dashboards.

Kimi Reviews

Анонимный отзыв
trustpilot.com
I bought a full year of Kimi Pro specifically for professional coding work, expecting K3 to be a serious alternative to Claude/Codex. The model itself is actually good. K3 is clearly smarter than many cheaper coding models. Unfortunately, the subscription limits make that almost irrelevant for serious professional use. A single normal Claude Code task exhausted my entire 5-hour usage allowance. Here are the actual numbers: * 373.6k input tokens * 56.4k output tokens * 15.1M cache-read tokens * 98% of input served from cache * 69 requests * 31m 50s of actual API activity * roughly 2 hours of the wall-clock session were IDLE So this was not five hours of coding. It was not even two hours of continuous coding. It was roughly 32 minutes of actual model activity, and that was enough to exhaust the entire 5-hour quota. This was not a stress test, benchmark, or some deliberately wasteful experiment. It was a normal task on a real software project. For someone who works with coding agents throughout an 8-hour workday, this makes the plan completely unsuitable as a primary development tool. What makes this especially frustrating is that there is no clear way to understand the real capacity of the subscription before committing to an annual plan. Kimi advertises professional coding capabilities, large context windows and agentic development, but does not clearly tell you something as basic as: * how much actual usage a 5-hour allowance represents; * approximately how many tokens it contains; * how cache reads are charged against quota; * what a realistic heavy Claude Code session consumes; * how many such sessions a Pro subscriber can expect. Had I known that approximately 32 minutes of real model activity could consume the entire 5-hour allowance, I would never have purchased an annual subscription. That is the core problem: the model is good, but the subscription is practically unusable for my professional workload, and the limits are not transparent enough to discover that before paying. For occasional coding it may be perfectly adequate. For serious daily software engineering with Claude Code or other agentic workflows, I strongly recommend buying only one month first and measuring the quota consumption on your real workload before committing to an annual plan.
Анонимный отзыв
trustpilot.com
Fraud all over false advertisement, they will sell, trade, give all your info to third parties without your consent, how safe those servers be to Americans when they are in China, they advertise as free, better than chat gpt and gimini.
Анонимный отзыв
trustpilot.com
I am documenting a serious billing issue with Kimi / Moonshot AI. I purchased a one-month Kimi subscription for USD 39. Kimi support later confirmed to me in writing that my subscription had ended and that there should be no further charges. Despite that confirmation, repeated USD 39 payment attempts continued afterward. I contacted support multiple times, explicitly withdrew authorization for any further recurring charges, asked them to stop all Stripe payment retries, and provided the billing information they requested. I also deleted my Link account and removed the payment information stored there. Even after all of that, another USD 39 payment authorization attempt was submitted to my bank and rejected. Kimi has acknowledged that the issue is still occurring and says the matter has been escalated internally for investigation. I am preserving the original receipt, invoice, support correspondence, failed-payment notifications and bank records. For privacy and security reasons, I am not publishing my email address, phone number, address, card details, invoice number, receipt number or private Stripe links. I will update this review if and when Kimi / Moonshot AI fully resolves the issue and confirms that no further payment attempts will occur.
Анонимный отзыв
trustpilot.com
I didn't think Kimi was too bad at all, but now there is some credit BS which has rendered it useless, its been 3 days I cannot use it and it says I have to wait another 4 days to use it again! WTF? Ridiculous! 1 star now since this rubbish!
Анонимный отзыв
trustpilot.com
Unflexible, stupid,not eveb the simpliest task, ruxe robithate it
Анонимный отзыв
trustpilot.com
Don't waste ur money use claude instead, your 20$ plan will last you about one day. Definetly not worth the money better off using claude.

Reviews are collected from public sources and translated into the page's language.

Kimi Pricing & Plans

Paid

Plus

$19 /month
Paid

Pro

$39 /month
Paid

Max

$99 /month
Paid

Ultra

$199 /month
Free

Free access

Free
Paid

Premium

$600 /year
Contact for pricing

Custom

Contact for pricing
Pay as you go

Flexible pay-as-you-go

Pay as you go
  • 1,000,000 tokens

Prices are based on the provider's information and may change.

Kimi Feature Comparison

Compare key features, prices and capabilities with similar tools

Feature
Kimi
Current tool
Replit
Google Antigravity
Ollama
Pricing modelFreemiumFreemiumFreemiumFreemium
Minimum pricefrom $19/monthfrom $20/monthfrom $19.99/monthfrom $20/month
Free access Yes Yes Yes Yes
User rating 100% positive 100% positive 100% positive 100% positive
Current tool

Frequently Asked Questions about Kimi

How does Kimi's Swarm mode differ from standard conversational AI?

Standard chat models respond to questions sequentially in a single linear thread. In Swarm mode, Kimi operates as an orchestrator, distributing a single complex task across multiple sub-agents simultaneously. These agents independently search distinct sources, write and test modular components, and merge their findings into a cohesive, verified output.

Can Kimi deploy live web applications and configure databases?

Yes. Within the Build module, users can describe an application, provide visual mockups, or specify functional requirements. The system generates the frontend and backend architectures, connects relational database structures, and deploys the finished product to a live environment with a single click.

What subscription tiers and pricing models does Kimi offer?

Kimi operates on a freemium model. It provides a permanent free plan alongside several monthly subscription tiers: Plus ($19/month), Pro ($39/month), Max ($99/month), and Ultra ($199/month). Users can also opt for an annual Premium subscription at $600/year, request custom terms, or use a flexible pay-as-you-go model.

What payment methods are supported for Kimi subscriptions?

Kimi officially accepts bank cards, Apple Pay, Google Pay, and Alipay for purchasing subscriptions and usage credits.

How can I pay for Kimi?

Kimi payment methods: bank card, Apple Pay, Google Pay, and Alipay.

Similar tools

Replit
Freemium
Replit ▲ 2 ▼ 0

Builds and deploys full-stack web applications, backends, and databases directly from natural language prompts within a unified cloud workspace. An autonomous agent generates multi-file code, installs dependencies, and runs automated browser tests to detect and fix errors.

AI coding toolsAI agents
Google Antigravity
Freemium
Google Antigravity ▲ 2 ▼ 0

Delegates software engineering tasks to autonomous AI agents that can plan complex steps, modify files, and execute terminal commands. Its browser-in-the-loop automation allows agents to autonomously run local development servers, test web interfaces, and inspect console errors.

AI coding toolsAI agents
Ollama
Freemium
Ollama ▲ 2 ▼ 0

Runs and orchestrates open-weight AI models locally or in the cloud without requiring manual CUDA setups or library wrangling. Built for developers, it enables private on-device execution alongside drop-in compatibility with standard API protocols.

AI coding toolsAI agents
Taskade
Freemium
Taskade ▲ 2 ▼ 0

Generates fully deployed software systems, such as CRMs, client portals, and dashboards, from simple text descriptions. It combines relational databases, specialized AI agents, and automated workflows across multiple project views without touching code.

ProductivityAI agents
Happycapy
Freemium
Happycapy ▲ 2 ▼ 0

Provides a browser-based visual cloud desktop where autonomous AI agents write code, execute scripts, and browse the web to deliver finished projects. It allows users to build web applications, conduct research, and run continuous background automations without configuring local environments or API keys.

ProductivityAI agents
View all alternatives Kimi