Provider

Moonshot AI

Every Moonshot AI model UltraGPT can reach, kept in sync with the live catalog — pricing, context window and capabilities for each one.

9 models from Moonshot AI Up to 1.05M tokens of context
Open the appBrowse every provider

9

Models tracked

1

Providers

7

Reasoning models

6

See images (vision)

1,049K

Largest context window

0

Free to run

Directory

All Moonshot AI models.

Search and filter this provider's lineup — pricing per million tokens, context window, modalities, reasoning effort and independent benchmark scores where available.

9 of 9 models

Kimi K2 Instruct is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It is optimized for...

Tool callingStreaming
Context window131.1K tok
Max output98.3K tok
ReleasedJul 11, 2025
Knowledge cutoffDec 31, 2024
TextText
Input$0.57 / 1M tok
Output$2.3 / 1M tok

Design Arena — ranked #95 across 5 categories

Kimi K2 0905 is the September update of [Kimi K2 0711](moonshotai/kimi-k2). It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32...

Tool callingStreamingJSON mode
Context window262.1K tok
Max output98.3K tok
ReleasedSep 4, 2025
Knowledge cutoffDec 31, 2024
TextText
Input$0.6 / 1M tok
Output$2.5 / 1M tok

Design Arena — ranked #91 across 2 categories

Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...

ReasoningTool callingStreamingJSON mode
Context window262.1K tok
Max output98.3K tok
ReleasedNov 6, 2025
Knowledge cutoff
TextText
Input$0.6 / 1M tok
Output$2.5 / 1M tok
Cache read$0.15 / 1M tok
Coding21.0

Design Arena — ranked #91 across 1 categories

Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...

ReasoningTool callingVisionStreamingJSON mode
Context window262.1K tok
Max output235.9K tok
ReleasedJan 27, 2026
Knowledge cutoff
Text + ImageText
Input$0.45 / 1M tok
Output$2.25 / 1M tok
Cache read$0.07 / 1M tok
Coding46.8

Design Arena — ranked #8 across 13 categories

Kimi K2.6 is Moonshot AI's next-generation multimodal model, designed for long-horizon coding, coding-driven UI/UX generation, and multi-agent orchestration. It handles complex end-to-end coding tasks across Python, Rust, and Go, and...

ReasoningTool callingVisionStreamingJSON mode
Context window262.1K tok
Max output235.9K tok
ReleasedApr 21, 2026
Knowledge cutoffJan 1, 2025
Text + ImageText
Input$0.95 / 1M tok
Output$4 / 1M tok
Cache read$0.16 / 1M tok
Coding61.8
Agentic22.1

Design Arena — ranked #2 across 21 categories

MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long contexts. It uses a native multimodal mixture-of-experts...

ReasoningTool callingVisionStreamingJSON mode
Context window262.1K tok
Max output235.9K tok
ReleasedJun 12, 2026
Knowledge cutoffJan 1, 2025
Text + ImageText
Input$0.71 / 1M tok
Output$3.5 / 1M tok
Cache read$0.15 / 1M tok
Intelligence26.3
Coding60.8
Agentic22.5

Design Arena — ranked #7 across 16 categories

Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...

ReasoningTool callingVisionStreamingJSON mode
Context window1.05M tok
Max output943.7K tok
ReleasedJul 16, 2026
Knowledge cutoff
Text + Image + VideoText
Input$2.65 / 1M tok
Output$13.28 / 1M tok
Cache read$0.303 / 1M tok
Effort: max, high, low
Intelligence43.8
Coding76.2
Agentic50.6

Design Arena — ranked #1 across 15 categories

Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...

ReasoningTool callingVisionStreamingJSON mode
Context window1.05M tok
Max output943.7K tok
ReleasedJul 16, 2026
Knowledge cutoff
Text + Image + VideoText
Input$3 / 1M tok
Output$15 / 1M tok
Cache read$0.3 / 1M tok
Effort: max, high, low
Intelligence43.8
Coding76.2
Agentic50.6

Design Arena — ranked #1 across 15 categories

This model always redirects to the latest model in the Kimi family.

ReasoningTool callingVisionStreamingJSON mode
Context window1.05M tok
Max output943.7K tok
ReleasedApr 27, 2026
Knowledge cutoff
Text + Image + VideoText
Input$2.1 / 1M tok
Output$10.95 / 1M tok
Cache read$0.23 / 1M tok
Effort: max, high, low

Every Moonshot AI model above, one subscription away.

UltraGPT puts this whole directory behind a single login and one flat monthly price — no separate bill per provider.