Model comparison
Claude Fable 5 vs Gemma 3 4B
Claude Fable 5 is the stronger model overall, scoring 66.3 to 27.9 on the Noometry Index. Gemma 3 4B costs 400× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
GoogleSummary
- They share 20 benchmarks with published results for both. Claude Fable 5 scores higher in 9 categories and Gemma 3 4B in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Fable 5 leads 88.4 to 15.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for Claude Fable 5 and 7.5% for Gemma 3 4B.
- Gemma 3 4B is cheaper at $0.04 / $0.08 per million input/output tokens, against $10 / $50 for Claude Fable 5.
- Claude Fable 5 accepts more context: 1M tokens versus 131K.
- Gemma 3 4B has downloadable open weights; the other is API-only.
Side by side
| Claude Fable 5 | Gemma 3 4B | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 66.3 | 27.9 |
| Released | 2026-06-07 | 2025-03-12 |
| Weights | Proprietary | Open |
| Context window | 1M | 131K |
| Max output | 128K | 4K |
| Input $ / M tokens | $10 | $0.04 |
| Output $ / M tokens | $50 | $0.08 |
| Results tracked | 61 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Fable 5 leads
Claude Fable 5: 67.0 (#4), Gemma 3 4B: 35.3 (#214)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| LMArena Coding | 1519 | 1230 |
| DeepSWE | 69.9% | — |
| FrontierCode | 53.5% | — |
| LMArena WebDev | 1626 | — |
| FrontierSWE | 47% | — |
| SciCode | 61% | — |
| WeirdML | 91.9% | — |
| MirrorCode | 63.9% | — |
| ALE-Bench | 2,041 | — |
Agentic & Tool Use Claude Fable 5 leads
Claude Fable 5: 54.1 (#2), Gemma 3 4B: 20.7 (#134)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| APEX-Agents | 63.6% | — |
| Berkeley Function Calling Leaderboard | — | 19.6% |
| Remote Labor Index | 16.1% | — |
| τ²-bench Banking | 39.7% | — |
| PostTrainBench | 41.8% | — |
| GBAEval | 74.5% | — |
| GDP.pdf | 30% | — |
| LMArena Search | 1230 | — |
| Vending-Bench 2 | 5,680 | — |
Reasoning Claude Fable 5 leads
Claude Fable 5: 76.8 (#4), Gemma 3 4B: 12.7 (#330)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| Kagi LLM Benchmark | 91.4% | 25.2% |
| Chess Puzzles | 41% | 0% |
| LMArena Hard Prompts | 1508 | 1253 |
| DTBench | 98.4% | 50.9% |
| LMCA | 61.1% | 2.8% |
| Epoch Capabilities Index | 163.6 | 116.02 |
| ARC-AGI-2 | 89.2% | — |
| SimpleBench | 81.9% | — |
| NYT Connections (extended) | 92.7% | — |
| ARC-AGI-1 | 98.5% | — |
| CritPt | 28.6% | — |
| EnigmaEval | 39.3% | — |
| EBR-Bench | 39.5% | — |
| Mystery Game Puzzles | 52% | — |
| Surface Evolver Bench | 95% | — |
| Bench to the Future 3 | 0.13 | — |
Math Claude Fable 5 leads
Claude Fable 5: 88.4 (#5), Gemma 3 4B: 15.8 (#286)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 7.5% |
| LMArena Math | 1519 | 1239 |
| FrontierMath (Tiers 1-3) | 87% | — |
| FrontierMath Tier 4 | 90.2% | — |
| ProofBench | 95% | — |
| FrontierMath Erdős | 0% | — |
Knowledge Claude Fable 5 leads
Claude Fable 5: 62.3 (#22), Gemma 3 4B: 11.8 (#294)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| GPQA Diamond | 85.9% | 23.2% |
| LMArena Expert | 1534 | 1223 |
| SimpleQA Verified | 70.7% | — |
| Vectara Hallucination Rate | — | 6.4% |
Multimodal Not comparable
Claude Fable 5: 45.6 (#14), Gemma 3 4B: —
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| LMArena Vision | 1326 | — |
| Blueprint-Bench 2 | 38.6% | — |
| Furniture Assembly | 35.8% | — |
| LMArena Document | 1496 | — |
Multilingual Claude Fable 5 leads
Claude Fable 5: 57.6 (#8), Gemma 3 4B: 42.7 (#189)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| LMArena Non-English | 1483 | 1273 |
| LMArena German | 1486 | 1281 |
| LMArena Russian | 1507 | 1294 |
| LMArena Chinese | 1546 | — |
| LMArena French | 1509 | — |
| LMArena Japanese | 1507 | — |
| LMArena Korean | 1493 | — |
| LMArena Spanish | 1503 | — |
Instruction Following Claude Fable 5 leads
Claude Fable 5: 78.8 (#7), Gemma 3 4B: 65.4 (#219)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| LMArena Instruction Following | 1503 | 1239 |
Long Context Claude Fable 5 leads
Claude Fable 5: 46.3 (#22), Gemma 3 4B: 38.6 (#189)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| LMArena Longer Query | 1510 | 1274 |
Writing & Preference Claude Fable 5 leads
Claude Fable 5: 76.1 (#5), Gemma 3 4B: 42.2 (#234)
| Benchmark | Claude Fable 5 | Gemma 3 4B |
|---|---|---|
| LMArena Text | 1491 | 1291 |
| LMArena Creative Writing | 1494 | 1270 |
| EQ-Bench Creative Writing | 1943 | 1068 |
| LMArena Multi-Turn | 1505 | 1255 |
| EQ-Bench 4 | 1340 | — |
Frequently asked questions
Is Claude Fable 5 better than Gemma 3 4B?
Claude Fable 5 is the stronger model overall, scoring 66.3 to 27.9 on the Noometry Index. Gemma 3 4B costs 400× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.
Which is cheaper, Claude Fable 5 or Gemma 3 4B?
Gemma 3 4B is cheaper. It lists at $0.04 per million input tokens and $0.08 per million output tokens; Claude Fable 5 lists at $10 and $50.
Is Claude Fable 5 or Gemma 3 4B better for coding?
Claude Fable 5 scores higher on coding benchmarks: 67.0 versus 35.3 in the Noometry coding category.
Which has the bigger context window?
Claude Fable 5 does, with 1M tokens against 131K.
How many benchmarks do Claude Fable 5 and Gemma 3 4B share?
20 benchmarks have published results for both models. Claude Fable 5 has 61 scored results on Noometry and Gemma 3 4B has 22.