The AI arena is free today

Open Superagent

Gemma 3 4B vs Hy4 flash

Hy4 flash leads the LLM Stats Score 40.8 to -2.4. Gemma 3 4B is 4.9x cheaper per token.

Google · Tencent · Updated for 2026

Which is better?

Hy4 flash leads the overall LLM Stats Score 40.8 to -2.4, ranking #84 overall.

On price, Gemma 3 4B is roughly 4.9x cheaper per token on a blended 3:1 input/output basis, which adds up quickly at production volume.

Hy4 flash also accepts a larger context window (960,000 input tokens), making it the stronger choice for long documents and large codebases.

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

Choose Gemma 3 4B

  • cost matters — it's about 4.9x cheaper per token
  • you need open weights you can self-host or fine-tune

Choose Hy4 flash

  • overall performance matters — it scores 40.8 and ranks #84 on LLM Stats
  • your work emphasizes reasoning and coding — it leads those capability indexes
  • you process long inputs — it offers a 960,000 token context window
  • you want the most recent training data — it shipped Nov 2026

At a glance

The differences that matter most.

Core performance indexes
-2.4
#388
40.8
#84
-3.5
#384
40.4
#77
-4.5
#301
23.5
#107
Cost, coverage & limits
Benchmark wins
—
—
Input price
$0.02 / M
$0.07 / M
Output price
$0.04 / M
$0.28 / M
Context window
131,072
960,000

Capability indexes

Additional strengths measured across groups of related public benchmarks

2 shared
Index
Gemma 3 4B
Hy4 flash
-11.8#247
21.1#90
-9.6#199
21.9#85
Conservative TrueSkill rating · higher is betterHow scores work

Individual benchmarks

26 reported for Gemma 3 4B · 9 for Hy4 flash

No common benchmarks found

Gemma 3 4B and Hy4 flashdon't have any common benchmark datasets to compare. They may have been evaluated on different testing suites.

Human preference

Blind head-to-head votes and playground preference scores

Pricing Analysis

Price comparison per million tokens

Gemma 3 4B costs less

For input processing, Gemma 3 4B ($0.02/1M tokens) is 3.5x cheaper than Hy4 flash ($0.07/1M tokens).

For output processing, Gemma 3 4B ($0.04/1M tokens) is 7.0x cheaper than Hy4 flash ($0.28/1M tokens).

In conclusion, Hy4 flash is more expensive than Gemma 3 4B.*

* Using a 3:1 ratio of input to output tokens

Lowest available price from all providers
Tue Nov 17 2026 • llm-stats.com
Google
Gemma 3 4B
Input tokens$0.02
Output tokens$0.04
Best providerDeepinfra
Tencent
Hy4 flash
Input tokens$0.07
Output tokens$0.28
Best providerTencent
Notice missing or incorrect data?

Model Size

Parameter count comparison

220.0B diff

Hy4 flash has 220.0B more parameters than Gemma 3 4B, making it 5500.0% larger.

Google
Gemma 3 4B
4.0Bparameters
Tencent
Hy4 flash
224.0Bparameters
4.0B
Gemma 3 4B
224.0B
Hy4 flash

Context Window

Maximum input and output token capacity

Hy4 flash accepts 960,000 input tokens compared to Gemma 3 4B's 131,072 tokens. Gemma 3 4B can generate longer responses up to 131,072 tokens, while Hy4 flash is limited to 64,000 tokens.

Google
Gemma 3 4B
Input131,072 tokens
Output131,072 tokens
Tencent
Hy4 flash
Input960,000 tokens
Output64,000 tokens
Tue Nov 17 2026 • llm-stats.com

Input capabilities

Documented input modalities across available providers

Both Gemma 3 4B and Hy4 flash support multimodal inputs.

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

Gemma 3 4B

Text
Images
Audio
Video

Hy4 flash

Text
Images
Audio
Video

Release Timeline

When each model was launched

Gemma 3 4B was released on 2025-03-12, while Hy4 flash was released on 2026-11-16.

Hy4 flash is 20 months newer than Gemma 3 4B.

Gemma 3 4B

Mar 12, 2025

1.7 years ago

Hy4 flash

Nov 16, 2026

1 days ago

1.7yr newer

Knowledge Cutoff

When training data ends

Gemma 3 4B has a documented knowledge cutoff of 2024-08-01, while Hy4 flash's cutoff date is not specified.

We can confirm Gemma 3 4B's training data extends to 2024-08-01, but cannot make a direct comparison without Hy4 flash's cutoff date.

Gemma 3 4B

Aug 2024

Hy4 flash

—

Provider Availability

Gemma 3 4B is available from DeepInfra. Hy4 flash is available from Tencent Cloud.

Gemma 3 4B

deepinfra logo
Deepinfra
Input Price:Input: $0.02/1MOutput Price:Output: $0.04/1M

Hy4 flash

tencent logo
Tencent
Input Price:Input: $0.07/1MOutput Price:Output: $0.28/1M
* Prices shown are per million tokens

Outputs Comparison

Notice missing or incorrect data?

Judge for yourself.

Run your own prompts against Gemma 3 4B and Hy4 flash side-by-side, then vote on the output you prefer.

Gemma 3 4B
✓ Preferred
Hy4 flash
Open in Playground

FAQ

Common questions about Gemma 3 4B vs Hy4 flash.

Which is better, Gemma 3 4B or Hy4 flash?

Hy4 flash leads the LLM Stats Score 40.8 to -2.4. Gemma 3 4B is made by Google and Hy4 flash is made by Tencent. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does Gemma 3 4B compare to Hy4 flash in benchmarks?

Gemma 3 4B scores IFEval: 90.2%, GSM8k: 89.2%, DocVQA: 75.8%, MATH: 75.6%, AI2D: 74.8%. Hy4 flash scores MMMU-Pro: 75.8%, OSWorld 2.1 (partial): 68.1%, DeepSWE 1.1: 64.6%, APEX-Agents: 63.1%, Humanity's Last Exam (with tools): 55.2%.

Is Gemma 3 4B cheaper than Hy4 flash?

Gemma 3 4B is 3.5x cheaper for input tokens. Gemma 3 4B costs $0.02/M input and $0.04/M output via deepinfra. Hy4 flash costs $0.07/M input and $0.28/M output via tencent.

What are the context window sizes for Gemma 3 4B and Hy4 flash?

Gemma 3 4B supports 131K tokens and Hy4 flash supports 960K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between Gemma 3 4B and Hy4 flash?

Key differences include LLM Stats Score (-2.4 vs 40.8), context window (131K vs 960K), input pricing ($0.02 vs $0.07/M), licensing (Gemma vs Unknown). See the full comparison above for benchmark-by-benchmark results.

Who makes Gemma 3 4B and Hy4 flash?

Gemma 3 4B is developed by Google and Hy4 flash is developed by Tencent.