Model comparison
Gemma 3 27B vs Sonar
Sonar is the stronger model overall, scoring 38.5 to 30.8 on the Noometry Index. Gemma 3 27B costs 10× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Last verified . 7 shared benchmarks.
Google
PerplexitySummary
- They share 7 benchmarks with published results for both. Gemma 3 27B scores higher in 0 categories and Sonar in 5 categories; 3 gaps are clear of the uncertainty.
- The widest gap is in coding, where Sonar leads 35.7 to 22.5.
- The biggest single-benchmark swing is LiveBench Math: 55.4% for Gemma 3 27B and 41.6% for Sonar.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $1 / $1 for Sonar.
- Gemma 3 27B accepts more context: 131K tokens versus 128K.
- Gemma 3 27B has downloadable open weights; the other is API-only.
Side by side
| Gemma 3 27B | Sonar | |
|---|---|---|
| Provider | Perplexity | |
| Noometry Index | 30.8 | 38.5 |
| Released | 2025-03-11 | 2024-01-01 |
| Weights | Open | Proprietary |
| Context window | 131K | 128K |
| Max output | 8K | 4K |
| Input $ / M tokens | $0.08 | $1 |
| Output $ / M tokens | $0.16 | $1 |
| Results tracked | 43 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Sonar leads
Gemma 3 27B: 22.5 (#334), Sonar: 35.7 (#221)
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LiveBench Coding | 39.9% | 35.1% |
| Aider Polyglot | 4.9% | — |
| SciCode | 21.2% | — |
| LMArena Coding | 1322 | — |
Agentic & Tool Use Not comparable
Gemma 3 27B: 25.1 (#110), Sonar: —
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
Reasoning Sonar leads
Gemma 3 27B: 16.7 (#301), Sonar: 21.1 (#227)
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LiveBench Reasoning | 43.8% | 46.3% |
| LiveBench Data Analysis | 51.5% | 37.9% |
| LiveBench | 50% | 46.9% |
| Kagi LLM Benchmark | 40.4% | — |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| LMArena Hard Prompts | 1340 | — |
| DTBench | 52.5% | — |
| LMCA | 12.3% | — |
| Epoch Capabilities Index | 130.04 | — |
Math Sonar leads
Gemma 3 27B: 25.9 (#265), Sonar: 33.7 (#200)
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LiveBench Math | 55.4% | 41.6% |
| OTIS Mock AIME 2024-2025 | 22.5% | — |
| LMArena Math | 1312 | — |
| MATH Level 5 | 74% | — |
Knowledge Not comparable
Gemma 3 27B: 25.5 (#261), Sonar: —
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| GPQA Diamond | 47.7% | — |
| Confabulations | 40.3% | — |
| Vectara Hallucination Rate | 7.4% | — |
| LMArena Expert | 1304 | — |
Multimodal Not comparable
Gemma 3 27B: 32.6 (#100), Sonar: —
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LMArena Vision | 1164 | — |
| GeoBench | 52% | — |
Multilingual Not comparable
Gemma 3 27B: 46.9 (#155), Sonar: —
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LMArena Non-English | 1334 | — |
| LMArena Chinese | 1346 | — |
| LMArena French | 1368 | — |
| LMArena German | 1362 | — |
| LMArena Japanese | 1287 | — |
| LMArena Korean | 1308 | — |
| LMArena Russian | 1349 | — |
| LMArena Spanish | 1349 | — |
Instruction Following Too close to call
Gemma 3 27B: 70.6 (#160), Sonar: 71.4 (#150)
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LiveBench Instruction Following | 74.9% | 76.2% |
| LMArena Instruction Following | 1321 | — |
Long Context Not comparable
Gemma 3 27B: 27.6 (#293), Sonar: —
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| Fiction.LiveBench | 33.3% | — |
| LMArena Longer Query | 1333 | — |
Writing & Preference Too close to call
Gemma 3 27B: 52.5 (#168), Sonar: 52.6 (#167)
| Benchmark | Gemma 3 27B | Sonar |
|---|---|---|
| LiveBench Language | 34.6% | 44.1% |
| LMArena Text | 1358 | — |
| LMArena Creative Writing | 1346 | — |
| Short-Story Creative Writing | 79.9% | — |
| EQ-Bench Creative Writing | 1266 | — |
| LMArena Multi-Turn | 1345 | — |
Frequently asked questions
Is Gemma 3 27B better than Sonar?
Sonar is the stronger model overall, scoring 38.5 to 30.8 on the Noometry Index. Gemma 3 27B costs 10× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Which is cheaper, Gemma 3 27B or Sonar?
Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; Sonar lists at $1 and $1.
Is Gemma 3 27B or Sonar better for coding?
Sonar scores higher on coding benchmarks: 35.7 versus 22.5 in the Noometry coding category.
Which has the bigger context window?
Gemma 3 27B does, with 131K tokens against 128K.
How many benchmarks do Gemma 3 27B and Sonar share?
7 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Sonar has 7.