Model Comparison

Claude Sonnet 4.5 vs Kimi K2-Instruct-0905

Claude Sonnet 4.5 significantly outperforms across most benchmarks.

Want to compare interactively?Try the playground

Performance Benchmarks

Comparative analysis across standard metrics

3 benchmarks

Claude Sonnet 4.5 outperforms in 3 benchmarks (AIME 2025, GPQA, Terminal-Bench), while Kimi K2-Instruct-0905 is better at 0 benchmarks.

Claude Sonnet 4.5 significantly outperforms across most benchmarks.

Wed Jun 03 2026 • llm-stats.com

Arena Performance

Human preference votes

Context Window

Maximum input and output token capacity

Only Claude Sonnet 4.5 specifies input context (200,000 tokens). Only Claude Sonnet 4.5 specifies output context (64,000 tokens).

Anthropic
Claude Sonnet 4.5
Input200,000 tokens
Output64,000 tokens
Moonshot AI
Kimi K2-Instruct-0905
Input- tokens
Output- tokens
Wed Jun 03 2026 • llm-stats.com

Input Capabilities

Supported data types and modalities

Claude Sonnet 4.5 supports multimodal inputs, whereas Kimi K2-Instruct-0905 does not.

Claude Sonnet 4.5 can handle both text and other forms of data like images, making it suitable for multimodal applications.

Claude Sonnet 4.5

Text
Images
Audio
Video

Kimi K2-Instruct-0905

Text
Images
Audio
Video

License

Usage and distribution terms

Claude Sonnet 4.5 is licensed under a proprietary license, while Kimi K2-Instruct-0905 uses MIT.

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

Claude Sonnet 4.5

Proprietary

Closed source

Kimi K2-Instruct-0905

MIT

Open weights

Release Timeline

When each model was launched

Claude Sonnet 4.5 was released on 2025-09-29, while Kimi K2-Instruct-0905 was released on 2025-09-05.

Claude Sonnet 4.5 is 1 month newer than Kimi K2-Instruct-0905.

Claude Sonnet 4.5

Sep 29, 2025

8 months ago

3w newer
Kimi K2-Instruct-0905

Sep 5, 2025

9 months ago

Knowledge Cutoff

When training data ends

Claude Sonnet 4.5 has a documented knowledge cutoff of 2025-01-31, while Kimi K2-Instruct-0905's cutoff date is not specified.

We can confirm Claude Sonnet 4.5's training data extends to 2025-01-31, but cannot make a direct comparison without Kimi K2-Instruct-0905's cutoff date.

Claude Sonnet 4.5

Jan 2025

Kimi K2-Instruct-0905

—

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion→

Key Takeaways

Anthropic
Larger context window (200,000 tokens)
Supports multimodal inputs
Higher AIME 2025 score (87.0% vs 49.5%)
Higher GPQA score (83.4% vs 75.1%)
Higher Terminal-Bench score (50.0% vs 25.0%)
Moonshot AI
Has open weights

Detailed Comparison

AI Model Comparison Table
Feature
Anthropic
Claude Sonnet 4.5
Moonshot AI
Kimi K2-Instruct-0905

FAQ

Common questions about Claude Sonnet 4.5 vs Kimi K2-Instruct-0905.

Which is better, Claude Sonnet 4.5 or Kimi K2-Instruct-0905?

Claude Sonnet 4.5 significantly outperforms across most benchmarks. Claude Sonnet 4.5 is made by Anthropic and Kimi K2-Instruct-0905 is made by Moonshot AI. The best choice depends on your use case — compare their benchmark scores, pricing, and capabilities above.

How does Claude Sonnet 4.5 compare to Kimi K2-Instruct-0905 in benchmarks?

Claude Sonnet 4.5 scores MMMLU: 89.1%, AIME 2025: 87.0%, TAU-bench Retail: 86.2%, GPQA: 83.4%, MMMUval: 77.8%. Kimi K2-Instruct-0905 scores MATH-500: 97.4%, MMLU-Redux: 92.7%, IFEval: 89.8%, AutoLogi: 89.5%, MMLU: 89.5%.

What are the context window sizes for Claude Sonnet 4.5 and Kimi K2-Instruct-0905?

Claude Sonnet 4.5 supports 200K tokens and Kimi K2-Instruct-0905 supports an unknown number of tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between Claude Sonnet 4.5 and Kimi K2-Instruct-0905?

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

Who makes Claude Sonnet 4.5 and Kimi K2-Instruct-0905?

Claude Sonnet 4.5 is developed by Anthropic and Kimi K2-Instruct-0905 is developed by Moonshot AI.