Kimi K2.5 vs Qwen3.5-122B-A10B Comparison

Comparing Kimi K2.5 and Qwen3.5-122B-A10B across benchmarks, pricing, and capabilities.

Want to compare interactively?Try the playground

Performance Benchmarks

Comparative analysis across standard metrics

23 benchmarks

Kimi K2.5 outperforms in 21 benchmarks (AA-LCR, BrowseComp, CharXiv-R, GPQA, HMMT 2025, Humanity's Last Exam, LiveCodeBench v6, LongBench v2, LVBench, MathVista-Mini, MMLU-Pro, MMMU-Pro, MMVU, OCRBench, Seal-0, SimpleVQA, SWE-Bench Verified, Terminal-Bench 2.0, VideoMMMU, WideSearch, ZEROBench), while Qwen3.5-122B-A10B is better at 2 benchmarks (MathVision, OmniDocBench 1.5).

Kimi K2.5 significantly outperforms across most benchmarks.

Tue Mar 17 2026 • llm-stats.com

Arena Performance

Human preference votes

Pricing Analysis

Price comparison per million tokens

Cost data unavailable.

Lowest available price from all providers

Tue Mar 17 2026 • llm-stats.com

Kimi K2.5

Input tokens$0.00

Output tokens$0.00

Best providerUnknown Organization

Qwen3.5-122B-A10B

Input tokens$0.40

Output tokens$3.20

Best providerNovita

Notice missing or incorrect data?Start an Issue→

Model Size

Parameter count comparison

878.0B diff

Kimi K2.5 has 878.0B more parameters than Qwen3.5-122B-A10B, making it 719.7% larger.

Kimi K2.5

1000.0Bparameters

Qwen3.5-122B-A10B

122.0Bparameters

1000.0B

Kimi K2.5

122.0B

Qwen3.5-122B-A10B

Context Window

Maximum input and output token capacity

Only Qwen3.5-122B-A10B specifies input context (262,144 tokens). Only Qwen3.5-122B-A10B specifies output context (64,000 tokens).

Kimi K2.5

Input- tokens

Output- tokens

Qwen3.5-122B-A10B

Input262,144 tokens

Output64,000 tokens

Tue Mar 17 2026 • llm-stats.com

Input Capabilities

Supported data types and modalities

Both Kimi K2.5 and Qwen3.5-122B-A10B support multimodal inputs.

They are both capable of processing various types of data, offering versatility in application.

Kimi K2.5

Text

Images

Audio

Video

Qwen3.5-122B-A10B

Text

Images

Audio

Video

License

Usage and distribution terms

Kimi K2.5 is licensed under MIT, while Qwen3.5-122B-A10B uses Apache 2.0.

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

Kimi K2.5

MIT

Open weights

Qwen3.5-122B-A10B

Apache 2.0

Open weights

Release Timeline

When each model was launched

Kimi K2.5 was released on 2026-01-27, while Qwen3.5-122B-A10B was released on 2026-02-24.

Qwen3.5-122B-A10B is 1 month newer than Kimi K2.5.

Kimi K2.5

Jan 27, 2026

1 months ago

Qwen3.5-122B-A10B

Feb 24, 2026

3 weeks ago

4w newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion→

Key Takeaways

Kimi K2.5

View details

Moonshot AI

Higher AA-LCR score (70.0% vs 66.9%)

Higher BrowseComp score (74.9% vs 63.8%)

Higher CharXiv-R score (77.5% vs 77.2%)

Higher GPQA score (87.6% vs 86.6%)

Higher HMMT 2025 score (95.4% vs 91.4%)

Higher Humanity's Last Exam score (50.2% vs 47.5%)

Higher LiveCodeBench v6 score (85.0% vs 78.9%)

Higher LongBench v2 score (61.0% vs 60.2%)

Higher LVBench score (75.9% vs 74.4%)

Higher MathVista-Mini score (90.1% vs 87.4%)

Higher MMLU-Pro score (87.1% vs 86.7%)

Higher MMMU-Pro score (78.5% vs 76.9%)

Higher MMVU score (80.4% vs 74.7%)

Higher OCRBench score (92.3% vs 92.1%)

Higher Seal-0 score (57.4% vs 44.1%)

Higher SimpleVQA score (71.2% vs 61.7%)

Higher SWE-Bench Verified score (76.8% vs 72.0%)

Higher Terminal-Bench 2.0 score (50.8% vs 49.4%)

Higher VideoMMMU score (86.6% vs 82.0%)

Higher WideSearch score (79.0% vs 60.5%)

Higher ZEROBench score (11.0% vs 9.0%)

Qwen3.5-122B-A10B

View details

Alibaba Cloud / Qwen Team

Larger context window (262,144 tokens)

Higher MathVision score (86.2% vs 84.2%)

Higher OmniDocBench 1.5 score (89.8% vs 88.8%)

Detailed Comparison

AI Model Comparison Table
Feature	Kimi K2.5	Qwen3.5-122B-A10B