Model comparison
gpt-oss-20b vs Sonar
Sonar is the stronger model overall, scoring 37.8 to 32.1 on the Noometry Index. gpt-oss-20b costs 28× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Last verified . 0 shared benchmarks.
PerplexitySummary
- The widest gap is in writing & preference, where Sonar leads 52.6 to 35.8.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $1 / $1 for Sonar.
- gpt-oss-20b accepts more context: 131K tokens versus 128K.
- gpt-oss-20b has downloadable open weights; the other is API-only.
Side by side
| gpt-oss-20b | Sonar | |
|---|---|---|
| Provider | OpenAI | Perplexity |
| Noometry Index | 32.1 | 37.8 |
| Released | 2025-08-05 | 2024-01-01 |
| Weights | Open | Proprietary |
| Context window | 131K | 128K |
| Max output | 16K | 4K |
| Input $ / M tokens | $0.018 | $1 |
| Output $ / M tokens | $0.09 | $1 |
| Results tracked | 34 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
gpt-oss-20b: 37.1 (#185), Sonar: 35.1 (#217)
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| SciCode | 34.4% | — |
| WeirdML | 40.9% | — |
| LiveBench Coding | — | 35.1% |
| LMArena Coding | 1307 | — |
| ALE-Bench | 566.05 | — |
Agentic & Tool Use Not comparable
gpt-oss-20b: 9.2 (#146), Sonar: —
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| Terminal-Bench | 3.4% | — |
Reasoning Sonar leads
gpt-oss-20b: 18.8 (#251), Sonar: 20.3 (#221)
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| Kagi LLM Benchmark | 53.2% | — |
| CritPt | 1.4% | — |
| Chess Puzzles | 4% | — |
| LiveBench Reasoning | — | 46.3% |
| LMArena Hard Prompts | 1274 | — |
| DTBench | 68% | — |
| LiveBench Data Analysis | — | 37.9% |
| LMCA | 14.5% | — |
| Epoch Capabilities Index | 137.81 | — |
| LiveBench | — | 46.9% |
Math gpt-oss-20b leads
gpt-oss-20b: 37.6 (#93), Sonar: 31.3 (#197)
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 65.3% | — |
| Omni-MATH | 56.5% | — |
| LiveBench Math | — | 41.6% |
| LMArena Math | 1317 | — |
Knowledge Not comparable
gpt-oss-20b: 34.6 (#186), Sonar: —
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| GPQA Diamond | 60.8% | — |
| MMLU-Pro | 74% | — |
| GPQA (HELM) | 59.4% | — |
| LMArena Expert | 1258 | — |
Multilingual Not comparable
gpt-oss-20b: 42.4 (#192), Sonar: —
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| LMArena Non-English | 1269 | — |
| LMArena Chinese | 1311 | — |
| LMArena German | 1249 | — |
| LMArena Japanese | 1245 | — |
| LMArena Korean | 1239 | — |
| LMArena Russian | 1279 | — |
| LMArena Spanish | 1267 | — |
Instruction Following Sonar leads
gpt-oss-20b: 62.0 (#235), Sonar: 71.5 (#146)
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| LiveBench Instruction Following | — | 76.2% |
| IFEval | 73.2% | — |
| LMArena Instruction Following | 1236 | — |
Long Context Not comparable
gpt-oss-20b: 37.9 (#203), Sonar: —
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| LMArena Longer Query | 1251 | — |
Writing & Preference Sonar leads
gpt-oss-20b: 35.8 (#260), Sonar: 52.6 (#166)
| Benchmark | gpt-oss-20b | Sonar |
|---|---|---|
| LMArena Text | 1288 | — |
| LMArena Creative Writing | 1201 | — |
| EQ-Bench Creative Writing | 666 | — |
| WildBench | 73.7% | — |
| LMArena Multi-Turn | 1268 | — |
| LiveBench Language | — | 44.1% |
Frequently asked questions
Is gpt-oss-20b better than Sonar?
Sonar is the stronger model overall, scoring 37.8 to 32.1 on the Noometry Index. gpt-oss-20b costs 28× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Which is cheaper, gpt-oss-20b or Sonar?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Sonar lists at $1 and $1.
Is gpt-oss-20b or Sonar better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.1 versus 35.1 in the Noometry coding category.
Which has the bigger context window?
gpt-oss-20b does, with 131K tokens against 128K.
How many benchmarks do gpt-oss-20b and Sonar share?
0 benchmarks have published results for both models. gpt-oss-20b has 34 scored results on Noometry and Sonar has 7.