Model Comparison

Hy3 vs Kimi K2 InstructWhich is better in 2026?

Hy3 significantly outperforms across most benchmarks.

Verdict: Hy3 vs Kimi K2 Instruct — which is better?

Hy3 (by Tencent) and Kimi K2 Instruct (by Moonshot AI) are two of the AI models people compare most. Here is how they stack up on benchmarks, price and capabilities, and which one to pick in 2026.

Hy3 outperforms in 2 benchmarks (GPQA, SWE-bench Multilingual), while Kimi K2 Instruct is better at 0 benchmarks. Hy3 significantly outperforms across most benchmarks.

Choose Hy3 if…

  • you want the strongest raw capability — it leads on 2 of 2 shared benchmarks
  • you want the most recent training data — it shipped Jul 2026

Choose Kimi K2 Instruct if…

  • you want predictable pricing at $0.50/M input and $0.50/M output

Performance Benchmarks

Comparative analysis across standard metrics

2 benchmarks

Hy3 outperforms in 2 benchmarks (GPQA, SWE-bench Multilingual), while Kimi K2 Instruct is better at 0 benchmarks.

Hy3 significantly outperforms across most benchmarks.

Tue Jul 21 2026 • llm-stats.com

Arena Performance

Human preference votes

Model Size

Parameter count comparison

705.0B diff

Kimi K2 Instruct has 705.0B more parameters than Hy3, making it 239.0% larger.

Tencent
Hy3
295.0Bparameters
Moonshot AI
Kimi K2 Instruct
1.0Tparameters
295.0B
Hy3
1000.0B
Kimi K2 Instruct

Context Window

Maximum input and output token capacity

Only Kimi K2 Instruct specifies input context (200,000 tokens). Only Kimi K2 Instruct specifies output context (200,000 tokens).

Tencent
Hy3
Input- tokens
Output- tokens
Moonshot AI
Kimi K2 Instruct
Input200,000 tokens
Output200,000 tokens
Tue Jul 21 2026 • llm-stats.com

License

Usage and distribution terms

Hy3 is licensed under Apache 2.0, while Kimi K2 Instruct uses MIT.

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

Hy3

Apache 2.0

Open weights

Kimi K2 Instruct

MIT

Open weights

Release Timeline

When each model was launched

Hy3 was released on 2026-07-06, while Kimi K2 Instruct was released on 2025-07-11.

Hy3 is 12 months newer than Kimi K2 Instruct.

Hy3

Jul 6, 2026

2 weeks ago

12mo newer
Kimi K2 Instruct

Jul 11, 2025

1.0 years ago

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

Tencent
Higher GPQA score (90.4% vs 75.1%)
Higher SWE-bench Multilingual score (75.8% vs 47.3%)
Moonshot AI
Larger context window (200,000 tokens)

Detailed Comparison

Interactive Arena

Judge for yourself.

Run your own prompts against Hy3 and Kimi K2 Instruct side-by-side, then vote on the output you prefer.

Hy3
✓ Preferred
Kimi K2 Instruct
Open in Playground
AI Model Comparison Table
Feature
Tencent
Hy3
Moonshot AI
Kimi K2 Instruct

FAQ

Common questions about Hy3 vs Kimi K2 Instruct.

Which is better, Hy3 or Kimi K2 Instruct?

Hy3 significantly outperforms across most benchmarks. Hy3 is made by Tencent and Kimi K2 Instruct is made by Moonshot AI. The best choice depends on your use case — compare their benchmark scores, pricing, and capabilities above.

How does Hy3 compare to Kimi K2 Instruct in benchmarks?

Hy3 scores DeepSearchQA: 91.0%, GPQA: 90.4%, IMO-AnswerBench: 90.0%, BrowseComp: 84.2%, MCP Atlas: 79.1%. Kimi K2 Instruct scores MATH-500: 97.4%, GSM8k: 97.3%, CBNSL: 95.6%, HumanEval: 93.3%, MMLU-Redux: 92.7%.

What are the context window sizes for Hy3 and Kimi K2 Instruct?

Hy3 supports an unknown number of tokens and Kimi K2 Instruct supports 200K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between Hy3 and Kimi K2 Instruct?

Key differences include licensing (Apache 2.0 vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes Hy3 and Kimi K2 Instruct?

Hy3 is developed by Tencent and Kimi K2 Instruct is developed by Moonshot AI.

Related comparisons