Moonshot AI
NativePrivateUpdated November 3, 2026Chinese foundation-model lab (Kimi)
- Website
- www.moonshot.ai
- CEO
- Yang Zhilin
- Founded
- 2023
- HQ
- Beijing, China
- Total Funding
- $8.5B
- Employees
- 201-500
Products & Offerings
Kimi K3
Flagship open-weight model with 2.8T parameters, built on Kimi Delta Attention, with native vision and a 1M-token context window.
Kimi API Platform
Paid developer API giving access to Kimi K3 and earlier Kimi models for building applications and agents.
Kimi (chatbot)
Consumer AI assistant on web, iOS and Android, offering long-context chat, search and multimodal input.
Kimi Agent (OK Computer)
General-purpose agent mode that plans, researches and produces websites, slides, dashboards and reports, with swarm-style sub-agents.
Kimi Code
Command-line coding agent powered by Kimi models, including the code-specialized Kimi K2.7 Code, for software development workflows.
Executive Summary
What they do
- Builds the Kimi family of open-weight frontier LLMs, led by 2.8T-parameter Kimi K3
- Runs the Kimi chatbot and the OK Computer agent for websites, slides and dashboards
- Sells paid API access to Kimi models and offers Kimi Code, a command-line coding agent
Why it matters
- Kimi K3 is reported as the largest open-weight model, with a 1M-token context window
- Kimi models reportedly underpin Cursor's Composer 2 and a Harvey legal model
- Raised $3.5B at $35B in July 2026; Hong Kong IPO targeted for early 2027
Best fit
- Teams self-hosting open-weight frontier models for coding and long-horizon agent workloads
- AI product builders fine-tuning or post-training on a strong open base model
- Organizations operating in China needing a leading domestic model provider
Watch-outs
- Anthropic alleges Kimi covertly routed user requests to Claude without user disclosure
- Beijing jurisdiction and state-fund backing complicate data residency and compliance reviews
- Hosted API data handling and model provenance need contractual and technical verification
Latest News
International Business Times UK · 4 days ago
NegativeAIBase · 5 days ago
PositiveWinBuzzer · 5 days ago
NegativeTOKENPOST · 5 days ago
NegativeUnlock Full Intelligence
Pro members get the complete briefing
- Complete analyst verdict
- Risk assessment
- All 6 Bluescore dimension scores
- Traction signals
- Bluescore history & momentum tracking
- Unlimited company access
Moonshot's models are among the strongest open-weight options available. Teams can still evaluate Kimi K3 weights on their own infrastructure, where Moonshot never sees the data. But Anthropic's September 2026 allegation is unresolved: it says Moonshot covertly sent Kimi user requests to Claude and passed the answers off as Kimi's. Combined with Chinese jurisdiction, that is a material risk for any enterprise using Moonshot's hosted app or API. Keep sensitive data off Moonshot-hosted services until the provenance and data-handling questions are answered, and watch the Hong Kong IPO filings for more disclosure.
Signal Provenance
Latest observation date and source for each signal category
| Signal Category | Last Updated | Source |
|---|---|---|
Funding Signals Last funding round, Funding stage | Oct 31, 2026 | Montclare research agents |
Hiring Velocity Headcount, Headcount growth | Oct 26, 2026 | Montclare research agents |
Technical Output Research papers, GitHub activity | Nov 3, 2026 | Montclare research agents |
News Sentiment News mentions, News sentiment | Nov 3, 2026 | Montclare research agents |
Partnership Signals Enterprise partnerships, Distribution partners | Nov 3, 2026 | Montclare research agents |
Revenue Indicators ARR estimate, Revenue growth | Nov 3, 2026 | Montclare research agents |
Leadership Signals CEO tenure, Founder-led | Nov 3, 2026 | Montclare research agents |
Competitive Position Category rank, Market share | Nov 3, 2026 | Montclare research agents |
Why this Bluescore?
6 dimensions, each scored 0–10 against all — companies. Each adds its weight × score to the Bluescore — see methodology