Moonshot AIKimi K2.7 Code

External Link

Kimi K2.7 Code is Moonshot AI's open-weight multimodal coding-agent model with 1T total parameters, 32B active parameters, and a 256K-token context window.

Token context
262,144 tokens
Inputs
Text, Image, Video
Outputs
Text
Released
Jun 11, 2026

Providers & Pricing

Estimate workload cost →
ProviderInput / 1MOutput / 1MCache read / 1MCache write / 1MObservedEvidence
$0.68$3.40$0.136Sep 3, 2026Provider-reported
$0.66$3.40$0.18Sep 3, 2026Provider-reported

At a Glance

Model Facts

Identity

Status
active
Developer ID
moonshotai/Kimi-K2.7-Code
Released
Jun 11, 2026
Version
Kimi K2.7 Code
License
modified-mit

Capacity

Token context
262,144 tokens
Maximum output
Unknown
Total parameters
1,000,000,000,000
Active parameters
32,000,000,000
Knowledge cutoff
Not reported

Interface and Access

Inputs
Text, Image, Video
Outputs
Text
API available
Yes
Open weights
Yes
Self-hostable
Yes
Reasoning
Yes
Vision
Yes
Tool calling
Yes
Capabilities
agents, chat, coding, reasoning, tools, vision

This page represents one developer model product. Dated API snapshots, serving endpoints, pricing, regions, quantizations, and service tiers attach as versioned aliases or provider details and never create duplicate public model records.

Version Lineage

PredecessorUnknown
SuccessorsUnknown
Aliaseskimi-k2.7-code, Kimi K2.7 Code, moonshotai/Kimi-K2.7-Code

Explicit Unknowns

  • Maximum output
  • Knowledge cutoff

Primary Evidence

Sources and Observation Date

Comparable Peers

Related Model Comparisons

All comparisons →
APairBContext
vsfamily variantsimage, text
vsfamily variantsimage, text
vscross-developer peerstext
vscross-developer peerstext
vscross-developer peersimage, text, video
vscross-developer peersimage, text, video
vscross-developer peersimage, text, video
vscross-developer peersimage, text, video
vscross-developer peersimage, text, video
vscross-developer peersimage, text, video
vsfamily variantsimage, text
vsfamily variantstext

Questions

Kimi K2.7 Code FAQs

What is Kimi K2.7 Code?+

Kimi K2.7 Code is Moonshot AI's open-weight multimodal coding-agent model with 1T total parameters, 32B active parameters, and a 256K-token context window. It is developed by Moonshot AI and its current sourced lifecycle status is active.

What inputs and outputs does Kimi K2.7 Code support?+

Kimi K2.7 Code's current record lists Text, Image, and Video as input and Text as output.

How much context does Kimi K2.7 Code support?+

The current record lists a 262,144-token context window and does not report a maximum output length.

Is Kimi K2.7 Code available through an API?+

Kimi K2.7 Code is marked as API-available. The directly sourced provider records currently include Deepinfra, Fireworks Ai, Openrouter, and Together Ai.

How large is Kimi K2.7 Code?+

The current record lists 1,000,000,000,000 total parameters and 32,000,000,000 active parameters. Its architecture is KimiK25ForConditionalGeneration.

What capabilities and tasks does Kimi K2.7 Code support?+

Kimi K2.7 Code's recorded capabilities are agents, chat, coding, reasoning, tools, and vision. Its supported tasks are coding, image, reasoning, text, and video.

How much does Kimi K2.7 Code cost through an API?+

The lowest directly sourced prices currently attached to Kimi K2.7 Code are $0.66 per million input tokens through Openrouter and $3.40 per million output tokens through Deepinfra. Prices are provider-specific and should be checked against each cited observation date.

Can Kimi K2.7 Code be self-hosted?+

Kimi K2.7 Code is an open-weight model. Self-hosting is supported, and the recorded license is modified-mit.

Does Kimi K2.7 Code have related versions or aliases?+

No predecessor is recorded. No successor is recorded. Recorded aliases are kimi-k2.7-code, Kimi K2.7 Code, and moonshotai/Kimi-K2.7-Code.

Which models can Kimi K2.7 Code be compared with?+

The current compatible peer set includes Kimi-K2.6, Kimi-K3, Qwen3 Coder Plus, DeepSeek-V4-Pro-0813, and Qwen3.8-Max. Compatibility requires shared sourced modalities and tasks; it does not imply that one model is better. Browse model comparisons

Send Feedback