AI Linkbase
Kimi logo
Freemiumchat

Kimi

Moonshot AI's long-context assistant, now powered by the Kimi K3 frontier model family.

Moonshot AI's assistant for long-context research, knowledge work, and coding. Its Kimi K3 model combines native multimodality, a 1M-token context window, and open weights for developers and teams evaluating frontier models.

4.4
Try Kimi Free →

Opens official website

long contextdocumentsresearch

Live change feed

Recent verified changes

Shared with the homepage and tool pages

newkimi

Kimi K3 launches as Moonshot's 2.8T open-weight frontier model

Moonshot AI's Kimi K3 combines native multimodality, a 1M-token context window, and open weights for long-horizon coding, knowledge work, and reasoning. Independent deployment, availability, and cost reviews are still in progress.

Verify official source →
verifiedLast verified 2026-07-24

Kimi K3 released — independent review in progress

Moonshot AI announced Kimi K3 on July 16, 2026. Its official materials describe a 2.8T-parameter, natively multimodal model with a 1M-token context window, designed for long-horizon coding, knowledge work, and reasoning.

Release stage
Released
Model scale
2.8T parameters (Moonshot claim)
Context window
1M tokens (Moonshot claim)
Model access
Open-weight; verify the current license and distribution terms before deployment
Editorial status
Strong launch; reliability, deployment cost, and independent benchmarks under review
Verify source: Moonshot AI official release

About Kimi

Kimi is Moonshot AI's general-purpose assistant for research, long-context work, coding, and knowledge tasks. The July 2026 Kimi K3 release is its most significant update to date: Moonshot describes K3 as a 2.8-trillion-parameter, natively multimodal open-weight model with a 1M-token context window. That makes Kimi a serious option for teams exploring long-horizon coding and open-model deployment, but launch benchmarks should not be treated as a universal buying verdict. Availability, regional access, operational reliability, licensing, hosting cost, and results on your own workloads still need to be evaluated before replacing an established primary assistant.

Pros

  • +Kimi K3 is positioned for long-horizon coding, knowledge work, and deep reasoning
  • +Moonshot lists native multimodality and a 1M-token context window for K3
  • +Excellent Chinese-English bilingual performance — best-in-class for cross-language workflows
  • +Real-time web search integrated directly into responses
  • +Multi-file analysis: PDF, Word, Excel, PowerPoint, images in one conversation
  • +Generous free tier with access to the full model — no watered-down free experience
  • +Strong performance on Chinese regulatory, legal, and academic content
  • +Fast, clean interface available on web, iOS, and Android

Cons

  • Headline benchmark claims still need independent, task-specific validation
  • Data privacy governed by Chinese regulations — may be unsuitable for sensitive enterprise data
  • Smaller international developer ecosystem and integration library
  • Advanced API access less documented than OpenAI or Anthropic equivalents
  • Open-weight deployment can carry material infrastructure and operating costs

Key Features

200,000-Character Long Context

Kimi's defining capability — upload an entire book, a full legal contract, a comprehensive annual report, or multiple lengthy research papers and analyze the complete content in a single conversation. The long-context window remains among the largest available in a consumer AI product in 2026.

Kimi K3 for Long-Horizon Work

Moonshot describes K3 as a model for long-horizon coding, knowledge work, and deep reasoning. Test it on representative repositories, documents, and review workflows rather than relying on a single public benchmark.

Real-Time Web Search

Kimi searches the web in real time to supplement its training data — accessing current news, recent research, up-to-date pricing, and live information. Particularly strong for Chinese-language web content, regulatory updates, and market information that international models may not index as thoroughly.

Multi-Format File Analysis

Upload PDFs, Word documents, Excel spreadsheets, PowerPoint presentations, and images — Kimi reads and analyzes all of them in a single conversation thread. Ideal for professionals who work across multiple document types and need AI that can synthesize information across formats.

Chinese-English Bilingual Intelligence

Kimi handles code-switching, translation, bilingual document analysis, and nuanced Chinese cultural and linguistic context with a quality that international models consistently underperform on. The preferred AI for professionals who operate across both language environments daily.

Mobile App with Voice Interaction

Kimi's iOS and Android apps bring the full model experience to mobile with voice input, making it accessible for on-the-go document queries, research, and writing assistance. The mobile experience is significantly more capable than many competing AI mobile apps in the Chinese market.

Pricing Plans

Free$0/ month
  • Current Kimi assistant access
  • Confirm current context and usage limits on the official site
  • Real-time web search
  • File uploads (PDF, Word, Excel)
  • Daily usage quota
ProCheck official site
  • Plan availability and limits vary by region and product surface
  • Confirm K3 access before purchasing
TeamCheck official site
  • All Pro features
  • Team workspace sharing
  • Admin management console
  • Usage analytics
  • Priority support
API / EnterpriseCheck official documentation
  • Moonshot API access
  • Custom context limits
  • Dedicated deployment options
  • Enterprise data agreements
  • SLA & dedicated support

Prices, plan limits, and promotions may change. Verify final pricing on the official site. Last verified: 2026-07-24.

🏆 Our Verdict

Kimi deserves a place on the shortlist for teams evaluating frontier open-weight models, especially for long-context coding, research, and knowledge work. K3's launch claims are meaningful enough to warrant immediate testing, but AI Linkbase's 4.4 rating remains deliberately conservative until reliability, cost, licensing, and independent performance evidence are clearer. Treat Kimi as a high-potential evaluation candidate—not as an automatic replacement for an established assistant.

Best For

Chinese-language professionals & researchersBilingual Chinese-English workflowsDocument-heavy analysis tasksStudents processing long academic papersLegal & compliance professionals in ChinaUsers needing free access to a capable long-context AI

Quick Info

Pricing
freemium
Category
chat
Rating
4.4 / 5.0
Last Reviewed
2026-07-24

Model access status

Kimi K3 · Released · risk Medium

Kimi K3
Frontier open weight model · Released
More chat tools