Model comparison
Deepseek Coder v2 vs Sonar
Sonar is the stronger model overall, scoring 37.8 to 35.3 on the Noometry Index.
Last verified . 0 shared benchmarks.
DeepSeek
PerplexitySummary
- The widest gap is in writing & preference, where Sonar leads 52.6 to 38.4.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.
Side by side
| Deepseek Coder v2 | Sonar | |
|---|---|---|
| Provider | DeepSeek | Perplexity |
| Noometry Index | 35.3 | 37.8 |
| Released | 2024-06-17 | 2024-01-01 |
| Weights | Open | Proprietary |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $1 |
| Output $ / M tokens | — | $1 |
| Results tracked | 24 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Deepseek Coder v2 leads
Deepseek Coder v2: 37.5 (#178), Sonar: 35.1 (#217)
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| BigCodeBench Instruct | 48.2% | — |
| LiveBench Coding | — | 35.1% |
| LMArena Coding | 1251 | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Reasoning Deepseek Coder v2 leads
Deepseek Coder v2: 22.6 (#175), Sonar: 20.3 (#221)
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LiveBench Reasoning | — | 46.3% |
| LMArena Hard Prompts | 1207 | — |
| LiveBench Data Analysis | — | 37.9% |
| LiveBench | — | 46.9% |
| WinoGrande | 83.7% | — |
Math Deepseek Coder v2 leads
Deepseek Coder v2: 32.3 (#188), Sonar: 31.3 (#197)
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LiveBench Math | — | 41.6% |
| LMArena Math | 1241 | — |
| GSM8K | 94.5% | — |
Knowledge Not comparable
Deepseek Coder v2: 31.7 (#209), Sonar: —
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LMArena Expert | 1181 | — |
| ARC (AI2) Challenge | 64.3% | — |
Multilingual Not comparable
Deepseek Coder v2: 36.4 (#235), Sonar: —
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LMArena Non-English | 1182 | — |
| LMArena Chinese | 1201 | — |
| LMArena French | 1184 | — |
| LMArena German | 1164 | — |
| LMArena Japanese | 1126 | — |
| LMArena Korean | 1104 | — |
| LMArena Russian | 1188 | — |
| LMArena Spanish | 1153 | — |
Instruction Following Sonar leads
Deepseek Coder v2: 61.9 (#237), Sonar: 71.5 (#146)
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LiveBench Instruction Following | — | 76.2% |
| LMArena Instruction Following | 1180 | — |
Long Context Not comparable
Deepseek Coder v2: 37.0 (#219), Sonar: —
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LMArena Longer Query | 1219 | — |
Writing & Preference Sonar leads
Deepseek Coder v2: 38.4 (#249), Sonar: 52.6 (#166)
| Benchmark | Deepseek Coder v2 | Sonar |
|---|---|---|
| LMArena Text | 1191 | — |
| LMArena Creative Writing | 1120 | — |
| LMArena Multi-Turn | 1176 | — |
| LiveBench Language | — | 44.1% |
Frequently asked questions
Is Deepseek Coder v2 better than Sonar?
Sonar is the stronger model overall, scoring 37.8 to 35.3 on the Noometry Index.
Is Deepseek Coder v2 or Sonar better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 37.5 versus 35.1 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and Sonar share?
0 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Sonar has 7.