Model comparison
Gemma 3 4B vs Hy3
Hy3 is the stronger model overall, scoring 44.2 to 28.1 on the Noometry Index. Gemma 3 4B costs 2.9× less per token, which makes it the better buy when Hy3's lead doesn't matter for your workload.
Last verified . 12 shared benchmarks.
Google
TencentSummary
- They share 12 benchmarks with published results for both. Gemma 3 4B scores higher in 0 categories and Hy3 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Hy3 leads 40.8 to 11.8.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $0.0825 / $0.33 for Hy3.
- Hy3 accepts more context: 262K tokens versus 131K.
Side by side
| Gemma 3 4B | Hy3 | |
|---|---|---|
| Provider | Tencent | |
| Noometry Index | 28.1 | 44.2 |
| Released | 2025-03-12 | 2026-07-06 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 4K | 128K |
| Input $ / M tokens | $0.04 | $0.0825 |
| Output $ / M tokens | $0.08 | $0.33 |
| Results tracked | 22 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Hy3 leads
Gemma 3 4B: 35.9 (#215), Hy3: 46.8 (#63)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Coding | 1230 | 1464 |
| LMArena WebDev | — | 1508 |
Agentic & Tool Use Not comparable
Gemma 3 4B: 20.9 (#142), Hy3: —
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 19.6% | — |
Reasoning Hy3 leads
Gemma 3 4B: 13.2 (#335), Hy3: 26.1 (#136)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Hard Prompts | 1253 | 1447 |
| Kagi LLM Benchmark | 25.2% | — |
| NYT Connections (extended) | — | 41.2% |
| Chess Puzzles | 0% | — |
| DTBench | 50.9% | — |
| LMCA | 2.8% | — |
| Epoch Capabilities Index | 116.02 | — |
Math Hy3 leads
Gemma 3 4B: 16.8 (#292), Hy3: 40.1 (#93)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Math | 1239 | 1475 |
| OTIS Mock AIME 2024-2025 | 7.5% | — |
Knowledge Hy3 leads
Gemma 3 4B: 11.8 (#299), Hy3: 40.8 (#114)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Expert | 1223 | 1460 |
| GPQA Diamond | 23.2% | — |
| Vectara Hallucination Rate | 6.4% | — |
Multilingual Hy3 leads
Gemma 3 4B: 42.5 (#194), Hy3: 53.5 (#65)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Non-English | 1273 | 1426 |
| LMArena German | 1281 | 1439 |
| LMArena Russian | 1294 | 1432 |
| LMArena Chinese | — | 1493 |
| LMArena French | — | 1461 |
| LMArena Japanese | — | 1392 |
| LMArena Korean | — | 1395 |
| LMArena Spanish | — | 1456 |
Instruction Following Hy3 leads
Gemma 3 4B: 65.2 (#225), Hy3: 75.1 (#70)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1239 | 1426 |
Long Context Hy3 leads
Gemma 3 4B: 38.7 (#194), Hy3: 44.1 (#75)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Longer Query | 1273 | 1442 |
Writing & Preference Hy3 leads
Gemma 3 4B: 42.0 (#239), Hy3: 62.2 (#81)
| Benchmark | Gemma 3 4B | Hy3 |
|---|---|---|
| LMArena Text | 1291 | 1439 |
| LMArena Creative Writing | 1271 | 1402 |
| LMArena Multi-Turn | 1255 | 1436 |
| EQ-Bench Creative Writing | 1068 | — |
Frequently asked questions
Is Gemma 3 4B better than Hy3?
Hy3 is the stronger model overall, scoring 44.2 to 28.1 on the Noometry Index. Gemma 3 4B costs 2.9× less per token, which makes it the better buy when Hy3's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 4B or Hy3?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; Hy3 lists at $0.0825 and $0.33.
Is Gemma 3 4B or Hy3 better for coding?
Hy3 scores higher on coding benchmarks: 46.8 versus 35.9 in the Noometry coding category.
Which has the bigger context window?
Hy3 does, with 262K tokens against 131K.
How many benchmarks do Gemma 3 4B and Hy3 share?
12 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Hy3 has 19.