Early build · scores computed by the deterministic engine from real, sourced developments (30 days) · weights v1.0-prior
Live
☰
Latest › Models › Moonshot AI (Kimi)
Vikshy Models · Moonshot AI (Kimi)

Moonshot AI (Kimi) models: pricing, context & what each is for

Moonshot AI (Kimi) has 12 models in the Vikshy catalogue, spanning Kimi K, Kimi K2, Kimi, Moonshot v1, compared side by side on context window, API pricing, modality, and what each is best for, every model linked to Moonshot AI (Kimi)'s official docs and pricing. Prices are per million tokens and can change, so the linked pricing page is the source of truth. Updated 2026-07-21.

ModelContextInput /MOutput /MModalityAccess
Kimi K3Frontier
Complex multi-step agents, deep research, and large-codebase work where the top tier matters.
1M
$3.00
$15.00
Text and vision
API (model id: kimi-k3) at api.moonshot.ai, plus kimi.com and the Kimi apps. Open weights planned on Hugging Face.
DocsPrice
Kimi K2.6Flagship
Teams wanting a capable open-weight agentic model with generous context at a moderate price.
256K
$0.95
$4.00
Text and vision
API (model id: kimi-k2.6) plus open weights on Hugging Face (moonshotai/Kimi-K2.6).
DocsPrice
Kimi K2 ThinkingReasoning
Math, planning, and step-by-step problems that benefit from deliberate reasoning.
256K
see pricing
see pricing
Text
API (model id: kimi-k2-thinking) plus open weights on Hugging Face (moonshotai/Kimi-K2-Thinking).
DocsPrice
kimi-thinking-previewReasoning
Harder multimodal reasoning tasks where a visible thinking pass helps.
128K
see pricing
see pricing
Text and vision
API only (model id: kimi-thinking-preview). Hosted preview.
DocsPrice
Kimi K2.5Balanced
Everyday agentic and coding work where you want good value on a proven open model.
256K
$0.60
$3.00
Text
API (model id: kimi-k2.5) plus open weights on Hugging Face (moonshotai/Kimi-K2.5).
DocsPrice
kimi-latestBalanced
Apps that want the newest Kimi without pinning a version, including vision inputs.
128K
see pricing
see pricing
Text and vision
API only (model id: kimi-latest). Hosted, no open weights for the alias.
DocsPrice
Kimi K2.7-CodeOpen
Agentic coding, refactors, and tool-driven development inside IDEs and CLI agents.
256K
$0.72
$3.50
Text
API (model id: kimi-k2.7-code) plus open weights on Hugging Face under a Modified MIT license.
DocsPrice
Kimi K2 (Instruct)Open
Self-hosting and research on a widely adopted open MoE base.
128K
Open
Open
Text
API (model id: kimi-k2-0711-preview) plus open weights on Hugging Face (moonshotai/Kimi-K2-Instruct).
DocsPrice
moonshot-v1-128kLegacy
Legacy integrations still calling the pre-K2 endpoints for long documents.
128K
$2.00
$5.00
Text
API only (model id: moonshot-v1-128k). Hosted.
DocsPrice
moonshot-v1-32kLegacy
Older apps needing moderate context on the legacy endpoint.
32K
$1.00
$3.00
Text
API only (model id: moonshot-v1-32k). Hosted.
DocsPrice
moonshot-v1-8kLegacy
Simple, cheap legacy calls that do not need long context.
8K
$0.20
$2.00
Text
API only (model id: moonshot-v1-8k). Hosted.
DocsPrice
moonshot-v1-128k-vision-previewLegacy
Older multimodal integrations built before the Kimi K vision models.
128K
$2.00
$5.00
Text and vision
API only (model id: moonshot-v1-128k-vision-preview). Hosted preview. Also offered in 8k and 32k vision variants.
DocsPrice
All models, compared →What is the Vikshy Score? →

Moonshot AI (Kimi) models, common questions

What is Moonshot AI (Kimi)'s most capable model?

Its highest-tier model in the catalogue is Kimi K3 (Frontier tier), built for complex multi-step agents, deep research, and large-codebase work where the top tier matters.

Which Moonshot AI (Kimi) model is cheapest?

By API input price, moonshot-v1-8k is the lowest at $0.20 per million input tokens ($2.00 output).

What is the largest context window Moonshot AI (Kimi) offers?

Kimi K3 has the largest listed context window, up to 1M tokens.

Does Moonshot AI (Kimi) offer open-weight models?

Yes. 7 of Moonshot AI (Kimi)'s catalogued models are available as open weights.