The AI arena is free today

Open Superagent

Gemini 2.5 Flash-Lite vs Kimi K2-Instruct-0905

Kimi K2-Instruct-0905 leads the LLM Stats Score 21.5 to 10.4.

Google · Moonshot AI · Updated for 2026

Which is better?

Kimi K2-Instruct-0905 leads the overall LLM Stats Score 21.5 to 10.4, ranking #202 overall.

In the 7 individual benchmarks reported for both models, Kimi K2-Instruct-0905 wins 5; 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 Gemini 2.5 Flash-Lite

  • you want predictable pricing at $0.10/M input and $0.40/M output

Choose Kimi K2-Instruct-0905

  • overall performance matters — it scores 21.5 and ranks #202 on LLM Stats
  • your work emphasizes reasoning and coding — it leads those capability indexes
  • you value its reported benchmark strengths — it wins 5 of 7 exact shared results
  • you want the most recent training data — it shipped Sep 2025

At a glance

The differences that matter most.

Gemini 2.5 Flash-Lite
Google
Kimi K2-Instruct-0905
Moonshot AI
Core performance indexes
LLM Stats Score
10.4
#280
21.5
#202
Reasoning
10.9
#274
21.8
#192
Coding
-2.9
#266
9.6
#184
Cost, coverage & limits
Benchmark wins
2 of 7
5 of 7
Input price
$0.10 / M
— / M
Output price
$0.40 / M
— / M
Context window
1,048,576
—

Capability indexes

Additional strengths measured across groups of related public benchmarks

1 shared
Index
Gemini 2.5 Flash-Lite
Kimi K2-Instruct-0905
Math
4.0#292
21.5#150
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

13 reported for Gemini 2.5 Flash-Lite · 29 for Kimi K2-Instruct-0905

7 shared

Gemini 2.5 Flash-Lite outperforms in 2 benchmarks (AIME 2025, Humanity's Last Exam), while Kimi K2-Instruct-0905 is better at 5 benchmarks (Aider-Polyglot, GPQA, LiveCodeBench, SimpleQA, SWE-Bench Verified).

Kimi K2-Instruct-0905 shows notably better performance in the majority of benchmarks.

Thu Oct 08 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Context Window

Maximum input and output token capacity

Only Gemini 2.5 Flash-Lite specifies input context (1,048,576 tokens). Only Gemini 2.5 Flash-Lite specifies output context (65,536 tokens).

Google
Gemini 2.5 Flash-Lite
Input1,048,576 tokens
Output65,536 tokens
Moonshot AI
Kimi K2-Instruct-0905
Input- tokens
Output- tokens
Thu Oct 08 2026 • llm-stats.com

Input capabilities

Documented input modalities across available providers

Gemini 2.5 Flash-Lite supports multimodal inputs, whereas Kimi K2-Instruct-0905 does not.

Gemini 2.5 Flash-Lite can handle both text and other forms of data like images, making it suitable for multimodal applications.

Gemini 2.5 Flash-Lite

Text
Images
Audio
Video

Kimi K2-Instruct-0905

Text
Images
Audio
Video

License

Usage and distribution terms

Gemini 2.5 Flash-Lite is licensed under Creative Commons Attribution 4.0 License, while Kimi K2-Instruct-0905 uses MIT.

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

Gemini 2.5 Flash-Lite

Creative Commons Attribution 4.0 License

Open weights

Kimi K2-Instruct-0905

MIT

Open weights

Release Timeline

When each model was launched

Gemini 2.5 Flash-Lite was released on 2025-06-17, while Kimi K2-Instruct-0905 was released on 2025-09-05.

Kimi K2-Instruct-0905 is 3 months newer than Gemini 2.5 Flash-Lite.

Gemini 2.5 Flash-Lite

Jun 17, 2025

1.3 years ago

Kimi K2-Instruct-0905

Sep 5, 2025

1.1 years ago

2mo newer

Knowledge Cutoff

When training data ends

Gemini 2.5 Flash-Lite has a documented knowledge cutoff of 2025-01-01, while Kimi K2-Instruct-0905's cutoff date is not specified.

We can confirm Gemini 2.5 Flash-Lite's training data extends to 2025-01-01, but cannot make a direct comparison without Kimi K2-Instruct-0905's cutoff date.

Gemini 2.5 Flash-Lite

Jan 2025

Kimi K2-Instruct-0905

—

Outputs Comparison

Notice missing or incorrect data?

Judge for yourself.

Run your own prompts against Gemini 2.5 Flash-Lite and Kimi K2-Instruct-0905 side-by-side, then vote on the output you prefer.

Gemini 2.5 Flash-Lite
✓ Preferred
Kimi K2-Instruct-0905
Open in Playground

FAQ

Common questions about Gemini 2.5 Flash-Lite vs Kimi K2-Instruct-0905.

Which is better, Gemini 2.5 Flash-Lite or Kimi K2-Instruct-0905?

Kimi K2-Instruct-0905 leads the LLM Stats Score 21.5 to 10.4. Gemini 2.5 Flash-Lite is made by Google and Kimi K2-Instruct-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 Gemini 2.5 Flash-Lite compare to Kimi K2-Instruct-0905 in benchmarks?

Gemini 2.5 Flash-Lite scores FACTS Grounding: 84.1%, Global-MMLU-Lite: 81.1%, MMMU: 72.9%, GPQA: 64.6%, Vibe-Eval: 51.3%. 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 Gemini 2.5 Flash-Lite and Kimi K2-Instruct-0905?

Gemini 2.5 Flash-Lite supports 1.0M 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 Gemini 2.5 Flash-Lite and Kimi K2-Instruct-0905?

Key differences include LLM Stats Score (10.4 vs 21.5), multimodal support (yes vs no), licensing (Creative Commons Attribution 4.0 License vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes Gemini 2.5 Flash-Lite and Kimi K2-Instruct-0905?

Gemini 2.5 Flash-Lite is developed by Google and Kimi K2-Instruct-0905 is developed by Moonshot AI.

Related comparisons