Private AI

Nano GPT logo
NanoGPT
Nano GPT logo
NanoGPT
Nano GPT logo
NanoGPT
Search
 
 
 
 
All conversations

Create

Chat
Studio

Explore

Models
Pricing
AI Apps
Blog

Library

Gallery
Workspace
Usage & billing
API
Suggestions and bugs
Use cases

|

Terms

|

Privacy

|

Refunds
Usage & billing
Teams
Referrals
Presets
Settings
FAQPress & mediaBrand assetsBenchmarksNano RPCEarn crypto
  1. Models
  2. Text
  3. Moonshot AI

Kimi K3

Kimi K3 is Moonshot AI's open-weight, always-thinking multimodal model for long-context reasoning, coding, tool use, and native image/video understanding.

  • Reasoning
  • Vision
  • Video Input
  • Tool Calling
  • Structured Output

Added Jul 16, 2026

Model weights

Pricing

Auto routing · per 1M tokens
Input
$0.45
Output
$9.60
Cache read
$0.13
Compare provider prices

Specifications

Context window
1M
Max output
1M

Benchmarks

Leaderboard results

BenchmarkScoreRank on NanoGPT
Epoch AI Benchmarking Hub

Epoch AI data, CC BY 4.0 · Updated Nov 14, 2026

Epoch Capabilities Index

158.6

#23 of 115

FrontierMath v2 (Tiers 1–3)

72.2%

#27 of 83

GPQA Diamond

93.1%

#22 of 119

SimpleQA Verified

50.6%

#29 of 82

FrontierMath v2 (Tier 4)

39.0%

#28 of 67

OTIS Mock AIME 2024–2025

97.2%

#35 of 121

MirrorCode

54.4%

#15 of 39

Chess Puzzles

37.0%

#40 of 82

Mystery Game Puzzles

26.0%

#39 of 77

Sierra τ-bench

tau2-bench data, MIT · Updated Aug 4, 2026

τ³-Banking

37.1%

#11 of 18

EQ-Bench

EQ-Bench data, MIT · Updated Nov 14, 2026

EQ-Bench 4

1319

#4 of 31

Creative Writing v3

2083

#6 of 96

Longform Writing

79.6

#18 of 97

Filtered and ranked to models on NanoGPT.

Providers

Choose or restrict providers with the API. Provider-selection prices include the selection markup. Provider selection guide · Pricing API (JSON, USD per 1K tokens)

Cache write /M
Auto
Auto (ZDR)
$0.45$9.60$0.13N/AN/AN/A
Auto
Auto (Anonymized)
$0.45$9.60$0.13N/AN/AN/A
Relace
Relace
FP4
UNK
$0.192$11.56$0.189N/AN/AN/A
Wafer
Wafer
MXFP4
US
$0.29$10.08$0.136N/AN/AN/A
Sail Research (Standard)
Sail Research (Standard)
FP4
GLB
$0.362$17.41$0.181N/AN/AN/A
Chutes
Chutes
MXFP4
US
$1.94$9.97$0.194N/AN/AN/A
EngyAI
EngyAI
GLB
$2.05$10.24$0.205N/AN/AN/A
Nube
Nube
$2.05$10.24$0.205N/AN/AN/A
io.net
io.net
FP4
US
$2.42$14.70$0.242N/AN/AN/A
DeepInfra
DeepInfra
MXFP4
US
$2.99$14.96$0.299N/AN/AN/A
RunAnywhere (Wally)
RunAnywhere (Wally)
MXFP4
US
$3.07$15.75$0.315N/AN/AN/A
BaseTen
BaseTen
FP8
US
$3.15$15.75$0.315N/AN/AN/A
Fireworks
Fireworks
US
$3.15$15.75$0.315N/AN/AN/A
Modal
Modal
MXFP4
US
$3.15$15.75$0.315N/AN/AN/A
Moonshot
Moonshot
SG
$3.15$15.75$0.315N/AN/AN/A
Novita
Novita
US
$3.15$15.75$0.315$3.15N/AN/A
Ollama
Ollama
UNK
$3.15$15.75$0.315N/AN/AN/A
Parasail
Parasail
FP4
US
$3.15$15.75$0.315N/AN/AN/A
Parasail Fast
Parasail Fast
FP4
US
$3.15$15.75$0.315N/AN/AN/A
Runware
Runware
UNK
$3.15$15.75$0.315N/AN/AN/A
TensorX
TensorX
EU
$3.15$15.75$0.787N/AN/AN/A
Together
Together
US
$3.15$15.75$0.315N/AN/AN/A
SCX (Global)
SCX (Global)
FP8
GLB
$3.29$16.46$0.335N/AN/AN/A
Alibaba
Alibaba
SG
$3.62$18.11$0.362N/AN/AN/A
Fireworks Fast
Fireworks Fast
US
$4.73$23.63$0.472N/AN/AN/A
Privacy:ZDR, prompts not retainedNo training on promptsRetention and training unclear
Batch API
ProviderInputOutputCached input
Auto
Automoonshotai/kimi-k3
$1.50/M$7.50/M$0.15/M

Batch rows for this model use normal routing at this price. A row that sets provider costs that provider's standard API price.

Related text models

Compare Kimi K3 with similar models from the same provider or model family.

Kimi K3.1 Pro

moonshotai/kimi-k3.1-pro

Kimi K3.1 Pro is a multimodal reasoning model for coding, 3D creation, document analysis and multi-step tool use. It supports a 1M-token context window and text, image and video inputs. Reasoning is always enabled, with low, high and max effort levels.

Kimi K3.1 Pro Highspeed

moonshotai/kimi-k3.1-pro-highspeed

Kimi K3.1 Pro Highspeed is the faster variant of K3.1 Pro, a multimodal reasoning model for coding, 3D creation, document analysis and multi-step tool use. It supports a 1M-token context window and text, image and video inputs. Reasoning is always enabled, with low, high and max effort levels.

Kimi K2.7 Code High-Speed

moonshotai/kimi-k2.7-code-highspeed

Kimi K2.7 Code High-Speed is the accelerated coding-focused variant tuned for fast agentic software engineering. It targets roughly 180 tokens per second, with short-context responses reaching up to about 260 tokens per second for rapid coding iterations.

Kimi K2.7 Code

moonshotai/kimi-k2.7-code

Kimi K2.7 Code is Moonshot AI's coding-focused agentic model built for long-horizon software engineering workflows. It supports native image input, tool calling, and forced thinking mode; instant/non-thinking mode is not supported.

Kimi Latest

moonshotai/kimi-latest

Always points to the newest recommended Kimi model, currently Kimi K3.1 Pro. Reasoning is always enabled. Select Kimi K3 explicitly to retain its checkpoint and pricing; K3.1 Pro Highspeed is available separately.

Kimi K2.6

moonshotai/kimi-k2.6

Kimi K2.6 is an open-source, native multimodal agentic model built for long-horizon coding, coding-driven design, and large-scale task orchestration. It can turn simple prompts and visual inputs into production-ready interfaces and full-stack workflows, and is designed to coordinate complex multi-agent plans with thousands of steps across code, documents, and spreadsheets.