The AI arena is free today

Open Superagent

Grok-4 Heavy vs Kimi K2-Thinking-0905

Grok-4 Heavy leads the LLM Stats Score 40.7 to 36.0.

xAI · Moonshot AI · Updated for 2026

Which is better?

Grok-4 Heavy leads the overall LLM Stats Score 40.7 to 36.0, ranking #54 overall.

In the 3 individual benchmarks reported for both models, Kimi K2-Thinking-0905 wins 2; this is a narrower head-to-head signal than the composite indexes.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose Grok-4 Heavy

  • overall performance matters — it scores 40.7 and ranks #54 on LLM Stats

Choose Kimi K2-Thinking-0905

  • you value its reported benchmark strengths — it wins 2 of 3 exact shared results
  • you need open weights you can self-host or fine-tune

At a glance

The differences that matter most.

Core performance indexes
40.7
#54
36.0
#81
39.0
#58
36.2
#74
17.7
#115
21.0
#91
Cost, coverage & limits
Benchmark wins
1 of 3
2 of 3
Input price
— / M
$0.47 / M
Output price
— / M
$2.00 / M
Context window
262,144

Capability indexes

Additional strengths measured across groups of related public benchmarks

1 shared
Index
Grok-4 Heavy
Kimi K2-Thinking-0905
41.5#6
38.0#27
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

6 reported for Grok-4 Heavy · 21 for Kimi K2-Thinking-0905

3 shared

Grok-4 Heavy outperforms in 1 benchmarks (GPQA), while Kimi K2-Thinking-0905 is better at 1 benchmark (Humanity's Last Exam).

Both models are evenly matched across the benchmarks.

Mon Sep 21 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Context Window

Maximum input and output token capacity

Only Kimi K2-Thinking-0905 specifies input context (262,144 tokens). Only Kimi K2-Thinking-0905 specifies output context (262,144 tokens).

xAI
Grok-4 Heavy
Input- tokens
Output- tokens
Moonshot AI
Kimi K2-Thinking-0905
Input262,144 tokens
Output262,144 tokens
Mon Sep 21 2026 • llm-stats.com

Input capabilities

Documented input modalities across available providers

Grok-4 Heavy supports multimodal inputs, whereas Kimi K2-Thinking-0905 does not.

Grok-4 Heavy can handle both text and other forms of data like images, making it suitable for multimodal applications.

Grok-4 Heavy

Text
Images
Audio
Video

Kimi K2-Thinking-0905

Text
Images
Audio
Video

License

Usage and distribution terms

Grok-4 Heavy is licensed under a proprietary license, while Kimi K2-Thinking-0905 uses MIT.

License differences may affect how you can use these models in commercial or open-source projects.

Grok-4 Heavy

Proprietary

Closed source

Kimi K2-Thinking-0905

MIT

Open weights

Release Timeline

When each model was launched

Kimi K2-Thinking-0905 was released on 2025-09-05, while Grok-4 Heavy's release date is not specified.

We can confirm Kimi K2-Thinking-0905's release timeline, but cannot make a direct age comparison without Grok-4 Heavy's release date.

Grok-4 Heavy

Kimi K2-Thinking-0905

Sep 5, 2025

1.0 years ago

Knowledge Cutoff

When training data ends

Grok-4 Heavy has a documented knowledge cutoff of 2024-12-31, while Kimi K2-Thinking-0905's cutoff date is not specified.

We can confirm Grok-4 Heavy's training data extends to 2024-12-31, but cannot make a direct comparison without Kimi K2-Thinking-0905's cutoff date.

Grok-4 Heavy

Dec 2024

Kimi K2-Thinking-0905

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion

Judge for yourself.

Run your own prompts against Grok-4 Heavy and Kimi K2-Thinking-0905 side-by-side, then vote on the output you prefer.

Grok-4 Heavy
✓ Preferred
Kimi K2-Thinking-0905
Open in Playground

FAQ

Common questions about Grok-4 Heavy vs Kimi K2-Thinking-0905.

Which is better, Grok-4 Heavy or Kimi K2-Thinking-0905?

Grok-4 Heavy leads the LLM Stats Score 40.7 to 36.0. Grok-4 Heavy is made by xAI and Kimi K2-Thinking-0905 is made by Moonshot AI. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does Grok-4 Heavy compare to Kimi K2-Thinking-0905 in benchmarks?

Grok-4 Heavy scores AIME 2025: 100.0%, HMMT25: 96.7%, GPQA: 88.4%, LiveCodeBench: 79.4%, USAMO25: 61.9%. Kimi K2-Thinking-0905 scores AIME 2025: 100.0%, HMMT 2025: 97.5%, MMLU-Redux: 94.4%, FRAMES: 87.0%, MMLU-Pro: 84.6%.

What are the context window sizes for Grok-4 Heavy and Kimi K2-Thinking-0905?

Grok-4 Heavy supports an unknown number of tokens and Kimi K2-Thinking-0905 supports 262K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between Grok-4 Heavy and Kimi K2-Thinking-0905?

Key differences include LLM Stats Score (40.7 vs 36.0), multimodal support (yes vs no), licensing (Proprietary vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes Grok-4 Heavy and Kimi K2-Thinking-0905?

Grok-4 Heavy is developed by xAI and Kimi K2-Thinking-0905 is developed by Moonshot AI.