Model comparison
Gemma 3 4B vs Qwen3-Coder 480B-A35B Instruct
Qwen3-Coder 480B-A35B Instruct is the stronger model overall, scoring 38.1 to 28.1 on the Noometry Index. Gemma 3 4B costs 60× less per token, which makes it the better buy when Qwen3-Coder 480B-A35B Instruct's lead doesn't matter for your workload.
Last verified . 13 shared benchmarks.
Google
Alibaba (Qwen)Summary
- They share 13 benchmarks with published results for both. Gemma 3 4B scores higher in 1 category and Qwen3-Coder 480B-A35B Instruct in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3-Coder 480B-A35B Instruct leads 37.0 to 11.8.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 25.2% for Gemma 3 4B and 49.5% for Qwen3-Coder 480B-A35B Instruct.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $1.50 / $7.50 for Qwen3-Coder 480B-A35B Instruct.
- Qwen3-Coder 480B-A35B Instruct accepts more context: 262K tokens versus 131K.
Side by side
| Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct | |
|---|---|---|
| Provider | Alibaba (Qwen) | |
| Noometry Index | 28.1 | 38.1 |
| Released | 2025-03-12 | 2025-04 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 4K | 66K |
| Input $ / M tokens | $0.04 | $1.50 |
| Output $ / M tokens | $0.08 | $7.50 |
| Results tracked | 22 | 25 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemma 3 4B: 35.9 (#215), Qwen3-Coder 480B-A35B Instruct: 35.5 (#223)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Coding | 1230 | 1412 |
| SWE-bench Verified (bash only) | — | 55.4% |
| LMArena WebDev | — | 1275 |
| GSO | — | 4.9% |
| WeirdML | — | 41.2% |
| ALE-Bench | — | 461.45 |
| AlgoTune | — | 1.44 |
Agentic & Tool Use Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 20.9 (#142), Qwen3-Coder 480B-A35B Instruct: 23.9 (#123)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Terminal-Bench | — | 27.2% |
| Berkeley Function Calling Leaderboard | 19.6% | — |
Reasoning Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 13.2 (#335), Qwen3-Coder 480B-A35B Instruct: 25.5 (#149)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 25.2% | 49.5% |
| LMArena Hard Prompts | 1253 | 1372 |
| Chess Puzzles | 0% | — |
| DTBench | 50.9% | — |
| LMCA | 2.8% | — |
| Epoch Capabilities Index | 116.02 | — |
Math Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 16.8 (#292), Qwen3-Coder 480B-A35B Instruct: 37.6 (#150)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Math | 1239 | 1365 |
| OTIS Mock AIME 2024-2025 | 7.5% | — |
Knowledge Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 11.8 (#299), Qwen3-Coder 480B-A35B Instruct: 37.0 (#162)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Expert | 1223 | 1338 |
| GPQA Diamond | 23.2% | — |
| Vectara Hallucination Rate | 6.4% | — |
Multilingual Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 42.5 (#194), Qwen3-Coder 480B-A35B Instruct: 47.7 (#148)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Non-English | 1273 | 1346 |
| LMArena German | 1281 | 1325 |
| LMArena Russian | 1294 | 1366 |
| LMArena Chinese | — | 1357 |
| LMArena French | — | 1398 |
| LMArena Japanese | — | 1310 |
| LMArena Korean | — | 1305 |
| LMArena Spanish | — | 1360 |
Instruction Following Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 65.2 (#225), Qwen3-Coder 480B-A35B Instruct: 71.6 (#147)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Instruction Following | 1239 | 1355 |
Long Context Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 38.7 (#194), Qwen3-Coder 480B-A35B Instruct: 42.0 (#131)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Longer Query | 1273 | 1378 |
Writing & Preference Qwen3-Coder 480B-A35B Instruct leads
Gemma 3 4B: 42.0 (#239), Qwen3-Coder 480B-A35B Instruct: 55.3 (#147)
| Benchmark | Gemma 3 4B | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Text | 1291 | 1357 |
| LMArena Creative Writing | 1271 | 1333 |
| LMArena Multi-Turn | 1255 | 1365 |
| EQ-Bench Creative Writing | 1068 | — |
Frequently asked questions
Is Gemma 3 4B better than Qwen3-Coder 480B-A35B Instruct?
Qwen3-Coder 480B-A35B Instruct is the stronger model overall, scoring 38.1 to 28.1 on the Noometry Index. Gemma 3 4B costs 60× less per token, which makes it the better buy when Qwen3-Coder 480B-A35B Instruct's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 4B or Qwen3-Coder 480B-A35B Instruct?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; Qwen3-Coder 480B-A35B Instruct lists at $1.50 and $7.50.
Is Gemma 3 4B or Qwen3-Coder 480B-A35B Instruct better for coding?
They score almost the same on coding (35.9 vs 35.5); test both on your own repository before choosing.
Which has the bigger context window?
Qwen3-Coder 480B-A35B Instruct does, with 262K tokens against 131K.
How many benchmarks do Gemma 3 4B and Qwen3-Coder 480B-A35B Instruct share?
13 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Qwen3-Coder 480B-A35B Instruct has 25.