Model comparison
Gemma 3 27B vs GLM-4.6
GLM-4.6 is the stronger model overall, scoring 41.4 to 30.8 on the Noometry Index. Gemma 3 27B costs 10× less per token, which makes it the better buy when GLM-4.6's lead doesn't matter for your workload.
Last verified . 23 shared benchmarks.
GoogleSummary
- They share 23 benchmarks with published results for both. Gemma 3 27B scores higher in 0 categories and GLM-4.6 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where GLM-4.6 leads 40.1 to 22.5.
- The biggest single-benchmark swing is Berkeley Function Calling Leaderboard: 29.5% for Gemma 3 27B and 72.4% for GLM-4.6.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $0.60 / $2.20 for GLM-4.6.
- GLM-4.6 accepts more context: 205K tokens versus 131K.
Side by side
| Gemma 3 27B | GLM-4.6 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | |
| Noometry Index | 30.8 | 41.4 |
| Released | 2025-03-11 | 2025-09-30 |
| Weights | Open | Open |
| Context window | 131K | 205K |
| Max output | 8K | 131K |
| Input $ / M tokens | $0.08 | $0.60 |
| Output $ / M tokens | $0.16 | $2.20 |
| Results tracked | 43 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-4.6 leads
Gemma 3 27B: 22.5 (#334), GLM-4.6: 40.1 (#148)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| SciCode | 21.2% | 38.4% |
| LMArena Coding | 1322 | 1449 |
| SWE-bench Verified (bash only) | — | 55.4% |
| Aider Polyglot | 4.9% | — |
| LMArena WebDev | — | 1340 |
| LiveBench Coding | 39.9% | — |
| ALE-Bench | — | 340.82 |
Agentic & Tool Use GLM-4.6 leads
Gemma 3 27B: 25.1 (#110), GLM-4.6: 32.3 (#66)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | 72.4% |
| Terminal-Bench | — | 24.5% |
Reasoning GLM-4.6 leads
Gemma 3 27B: 16.7 (#301), GLM-4.6: 23.7 (#172)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 47.4% |
| CritPt | 0% | 1.1% |
| LMArena Hard Prompts | 1340 | 1440 |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | 43.8% | — |
| DTBench | 52.5% | — |
| LiveBench Data Analysis | 51.5% | — |
| LMCA | 12.3% | — |
| Epoch Capabilities Index | 130.04 | — |
| LiveBench | 50% | — |
Math GLM-4.6 leads
Gemma 3 27B: 25.9 (#265), GLM-4.6: 39.1 (#111)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| LMArena Math | 1312 | 1432 |
| OTIS Mock AIME 2024-2025 | 22.5% | — |
| LiveBench Math | 55.4% | — |
| MATH Level 5 | 74% | — |
| FrontierMath (Feb 2025 set) | — | 3.8% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GLM-4.6 leads
Gemma 3 27B: 25.5 (#261), GLM-4.6: 40.2 (#124)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| Vectara Hallucination Rate | 7.4% | 9.5% |
| LMArena Expert | 1304 | 1431 |
| GPQA Diamond | 47.7% | — |
| Confabulations | 40.3% | — |
Multimodal Not comparable
Gemma 3 27B: 32.6 (#100), GLM-4.6: —
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| LMArena Vision | 1164 | — |
| GeoBench | 52% | — |
Multilingual GLM-4.6 leads
Gemma 3 27B: 46.9 (#155), GLM-4.6: 53.5 (#66)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| LMArena Non-English | 1334 | 1426 |
| LMArena Chinese | 1346 | 1499 |
| LMArena French | 1368 | 1459 |
| LMArena German | 1362 | 1447 |
| LMArena Japanese | 1287 | 1393 |
| LMArena Korean | 1308 | 1400 |
| LMArena Russian | 1349 | 1419 |
| LMArena Spanish | 1349 | 1436 |
Instruction Following GLM-4.6 leads
Gemma 3 27B: 70.6 (#160), GLM-4.6: 74.3 (#98)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| LMArena Instruction Following | 1321 | 1410 |
| LiveBench Instruction Following | 74.9% | — |
Long Context GLM-4.6 leads
Gemma 3 27B: 27.6 (#293), GLM-4.6: 43.4 (#94)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| LMArena Longer Query | 1333 | 1422 |
| Fiction.LiveBench | 33.3% | — |
Writing & Preference GLM-4.6 leads
Gemma 3 27B: 52.5 (#168), GLM-4.6: 61.1 (#90)
| Benchmark | Gemma 3 27B | GLM-4.6 |
|---|---|---|
| LMArena Text | 1358 | 1440 |
| LMArena Creative Writing | 1346 | 1411 |
| EQ-Bench Creative Writing | 1266 | 1411 |
| LMArena Multi-Turn | 1345 | 1427 |
| Short-Story Creative Writing | 79.9% | — |
| LiveBench Language | 34.6% | — |
Frequently asked questions
Is Gemma 3 27B better than GLM-4.6?
GLM-4.6 is the stronger model overall, scoring 41.4 to 30.8 on the Noometry Index. Gemma 3 27B costs 10× less per token, which makes it the better buy when GLM-4.6's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 27B or GLM-4.6?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; GLM-4.6 lists at $0.60 and $2.20.
Is Gemma 3 27B or GLM-4.6 better for coding?
GLM-4.6 scores higher on coding benchmarks: 40.1 versus 22.5 in the Noometry coding category.
Which has the bigger context window?
GLM-4.6 does, with 205K tokens against 131K.
How many benchmarks do Gemma 3 27B and GLM-4.6 share?
23 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and GLM-4.6 has 29.