Moonshot AI · Language model
Kimi
Chinese AI lab (Moonshot AI) with the open-weight Kimi K2 family and an assistant that focuses strongly on agents, deep research and long documents.
1487#3 of 12 in the index
Elo, independently measuredQuality from blind duels on LMArena: users choose between two anonymous answers. A difference of around 20 points is noticeable; a few points is noise. How we measure
Elo, independently measuredQuality from blind duels on LMArena: users choose between two anonymous answers. A difference of around 20 points is noticeable; a few points is noise. How we measure
−1 Eloin 30 days
- Editorial scoreOur own assessment on 10 of suitability for professional work, price and privacy. This is a judgement, not a measurement; that is why it always sits alongside the Elo. How we measure
- 7,6
- GDPR profileOur summary of what the provider itself publishes about your data: EU hosting possible, no training on your data for business use, and self-hosting available. Caution applies to a model with a documented problem, such as an investigation by a regulator. How we measure
- GDPR: Caution
- Entry price
- Freepaid from $19 / month
- Context
- 1M tokens
- Usage on OpenRouter
- 3.79% of tokens
Quality over time
161 daily measurementsThe line follows the strongest variant of the family; the figure at the top is the measured flagship.
1490147714631450
19 March 20269 November 2026
Source: LMArena · retrieved 9 November 2026
Your data
GDPR: Caution–Hosting in the EU possible
–No training on your data (business)
✓Self-hostable (open weights)
The international service is provided by Moonshot AI Pte. Ltd. (Singapore); according to the privacy policy, data is stored on servers in Singapore and content may be used to improve the service. The K2 models are open-weight and can be self-hosted.
GDPR check by professionMoonshot AI's models
Kimi is a family. Choose the heaviest for difficult tasks, the fastest for volume.
| Model | Type | Elo | Price in | Price out | Context | Buzz |
|---|---|---|---|---|---|---|
| Kimi K3.1 Pro | Flagship (newest) | – | $2 | $10 | 1.048.576 | 49 |
| Kimi K3 | Flagship (previous) | – | $0.19 | $16 | 1.048.576 | 495 |
| Kimi K3 Max | Max reasoning | – | – | – | – | 10 |
| Kimi K2.6 | Flagship | – | $0.65 | $3.41 | 262.144 | 12 |
| Kimi K2.5 | Balanced | – | $0.53 | $3.32 | 262.144 | 15 |
| Kimi K2.7 Code | Coding model | – | $0.6 | $3.5 | 262.144 | 7 |
| Kimi K2 0905 | Snapshot | – | $0.6 | $2.5 | 262.144 | 1 |
Source: LMArena, OpenRouter, Hacker News · retrieved 9 November 2026
Subscriptions
Kimi (free)FreeBasic chat on kimi.com
Moderato$19 / monthK2.6, Deep Research, Kimi Code and agent credits; $15/month when paid annually
API, Kimi K3$3.00 / $15.00 per 1MNewest flagship; cached input $0.30 per 1M, via platform.kimi.ai
API, Kimi K2.6$0.95 / $4.00 per 1MBudget option; cached input $0.16 per 1M, via platform.kimi.ai
Where it excels
- 1Flagship K3 (2.8T parameters, 1M context) and open-weight K2.6 can be self-hosted
- 2Multimodal: text, image and video in one architecture
- 3Strong at agent work and deep research
- 4Affordable API with a substantial cache discount
Kimi by profession
4 professions where we recommend it, listed first.
MarketingEducationCustomer serviceCommunications & mediaLegal professionAccountancyHealthcareRecruitmentE-commerceReal estateLogisticsHospitalityConstruction & installationHR & personnelSales & prospectingGovernment & municipalitiesIndustry & manufacturingFinancial adviceVeterinary practices & animal carePhysiotherapyDentistryArchitecture & urban planningAutomotive sectorChildcareInsurance sectorSport & fitnessInstallation technologyAgricultural sectorTourism & travel sectorPharmacy & pharmaceuticalsNon-profit & fundraisingDebt collection & receivables managementNotaryStaffing & flexible workMental health & psychologyBeauty & wellnessGP practicesEvents sectorRenewable energy & the energy transitionIT & software developmentHousing associations & owners' associationsFacilities management & cleaningTranslation & localisationResearch & scienceBuilding management & installationsElderly care & long-term careRadiology & medical imagingBanks & fintechPathologyEye careProcurement & tenderingSocial domain & youth careOccupational health & absenteeismCorporate training & L&DGreenhouse horticultureHotels & holiday accommodationFashion & textilesFood industryPharmaceutical industryPhotographyWater managementCybersecurityGraphic design & designMusic & audio productionGame developmentInterior designVideo & film production
The full index on one scale
Ordered by quality, height is our editorial score. Kimi is featured.
#1Claude#2Gemini#3Grok#4ChatGPT#5Copilot#6Qwen#7Perplexity#8DeepSeek#9Kimi#10Le Chat (Mistral)#11GLM (Z.ai)#12Llama
Scale 0 to 10 · editorial score
1Gemini8,62Claude9,03Kimi7,64ChatGPT8,95Copilot8,36Qwen7,97GLM (Z.ai)7,78DeepSeek7,49Grok7,810Le Chat (Mistral)8,411Perplexity8,012Llama7,6
Compare Kimi with another
What the community says
Kimi K3.1 Pro and K3.1 Pro Highspeed1096 points · 681 commentsKimi K3.1 Pro model weights are out640 points · 191 commentsKimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed from four SSDs279 points · 156 commentsShow HN: Training a model to identify AI web content from structure alone74 points · 28 comments
Hacker News, last 90 days, praise and criticism.
