Model comparison
Gemma 3 4B vs Inkling
Inkling is the stronger model overall, scoring 44.1 to 28.1 on the Noometry Index. Gemma 3 4B costs 51× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
GoogleSummary
- They share 19 benchmarks with published results for both. Gemma 3 4B scores higher in 1 category and Inkling in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Inkling leads 55.1 to 11.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 7.5% for Gemma 3 4B and 88.9% for Inkling.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- Gemma 3 4B accepts more context: 131K tokens versus 66K.
Side by side
| Gemma 3 4B | Inkling | |
|---|---|---|
| Provider | Thinking Machines Lab | |
| Noometry Index | 28.1 | 44.1 |
| Released | 2025-03-12 | 2026-07-15 |
| Weights | Open | Open |
| Context window | 131K | 66K |
| Max output | 4K | 66K |
| Input $ / M tokens | $0.04 | $1.87 |
| Output $ / M tokens | $0.08 | $4.68 |
| Results tracked | 22 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 3 4B leads
Gemma 3 4B: 35.9 (#215), Inkling: 34.5 (#234)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| LMArena Coding | 1230 | 1464 |
| FrontierCode | — | 14% |
| LMArena WebDev | — | 1413 |
| FrontierSWE | — | 4.1% |
| SciCode | — | 47% |
| WeirdML | — | 32.3% |
| ALE-Bench | — | 946 |
Agentic & Tool Use Inkling leads
Gemma 3 4B: 20.9 (#142), Inkling: 29.6 (#85)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| Berkeley Function Calling Leaderboard | 19.6% | — |
| τ²-bench Banking | — | 25% |
Reasoning Inkling leads
Gemma 3 4B: 13.2 (#335), Inkling: 40.4 (#56)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| Chess Puzzles | 0% | 21% |
| LMArena Hard Prompts | 1253 | 1451 |
| DTBench | 50.9% | 87.5% |
| LMCA | 2.8% | 37.6% |
| Epoch Capabilities Index | 116.02 | 148.54 |
| ARC-AGI-2 | — | 36.5% |
| SimpleBench | — | 50% |
| Kagi LLM Benchmark | 25.2% | — |
| ARC-AGI-1 | — | 79.5% |
| CritPt | — | 5.4% |
Math Inkling leads
Gemma 3 4B: 16.8 (#292), Inkling: 31.3 (#225)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 7.5% | 88.9% |
| LMArena Math | 1239 | 1479 |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| FrontierMath Tier 4 | — | 4.9% |
| ProofBench | — | 0% |
Knowledge Inkling leads
Gemma 3 4B: 11.8 (#299), Inkling: 55.1 (#49)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| GPQA Diamond | 23.2% | 88.3% |
| LMArena Expert | 1223 | 1465 |
| SimpleQA Verified | — | 40.3% |
| Vectara Hallucination Rate | 6.4% | — |
Multilingual Inkling leads
Gemma 3 4B: 42.5 (#194), Inkling: 54.0 (#52)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| LMArena Non-English | 1273 | 1434 |
| LMArena German | 1281 | 1446 |
| LMArena Russian | 1294 | 1429 |
| LMArena Chinese | — | 1490 |
| LMArena French | — | 1458 |
| LMArena Japanese | — | 1429 |
| LMArena Korean | — | 1404 |
| LMArena Spanish | — | 1448 |
Instruction Following Inkling leads
Gemma 3 4B: 65.2 (#225), Inkling: 75.1 (#71)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| LMArena Instruction Following | 1239 | 1426 |
Long Context Inkling leads
Gemma 3 4B: 38.7 (#194), Inkling: 43.8 (#86)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| LMArena Longer Query | 1273 | 1434 |
Writing & Preference Inkling leads
Gemma 3 4B: 42.0 (#239), Inkling: 65.2 (#51)
| Benchmark | Gemma 3 4B | Inkling |
|---|---|---|
| LMArena Text | 1291 | 1441 |
| LMArena Creative Writing | 1271 | 1387 |
| EQ-Bench Creative Writing | 1068 | 1611 |
| LMArena Multi-Turn | 1255 | 1436 |
| EQ-Bench 4 | — | 1226 |
Frequently asked questions
Is Gemma 3 4B better than Inkling?
Inkling is the stronger model overall, scoring 44.1 to 28.1 on the Noometry Index. Gemma 3 4B costs 51× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 4B or Inkling?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; Inkling lists at $1.87 and $4.68.
Is Gemma 3 4B or Inkling better for coding?
Gemma 3 4B scores higher on coding benchmarks: 35.9 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
Gemma 3 4B does, with 131K tokens against 66K.
How many benchmarks do Gemma 3 4B and Inkling share?
19 benchmarks have published results for both models. Gemma 3 4B has 22 scored results on Noometry and Inkling has 41.