Model comparison
Qwen3 8B vs Sonar
Sonar is the stronger model overall, scoring 38.5 to 33.7 on the Noometry Index. Qwen3 8B costs 3.2× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Last verified . 0 shared benchmarks.
Alibaba (Qwen)
PerplexitySummary
- The widest gap is in reasoning, where Sonar leads 21.1 to 16.6.
- Qwen3 8B is cheaper at $0.18 / $0.70 per million input/output tokens, against $1 / $1 for Sonar.
- Qwen3 8B accepts more context: 131K tokens versus 128K.
- Qwen3 8B has downloadable open weights; the other is API-only.
Side by side
| Qwen3 8B | Sonar | |
|---|---|---|
| Provider | Alibaba (Qwen) | Perplexity |
| Noometry Index | 33.7 | 38.5 |
| Released | 2025-04 | 2024-01-01 |
| Weights | Open | Proprietary |
| Context window | 131K | 128K |
| Max output | 8K | 4K |
| Input $ / M tokens | $0.18 | $1 |
| Output $ / M tokens | $0.70 | $1 |
| Results tracked | 11 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Sonar leads
Qwen3 8B: 34.0 (#248), Sonar: 35.7 (#221)
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| SciCode | 22.6% | — |
| LiveBench Coding | — | 35.1% |
Agentic & Tool Use Not comparable
Qwen3 8B: 30.2 (#78), Sonar: —
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| Berkeley Function Calling Leaderboard | 42.6% | — |
Reasoning Sonar leads
Qwen3 8B: 16.6 (#303), Sonar: 21.1 (#227)
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| CritPt | 0% | — |
| Chess Puzzles | 5% | — |
| LiveBench Reasoning | — | 46.3% |
| DTBench | 59.7% | — |
| LiveBench Data Analysis | — | 37.9% |
| LMCA | 8.8% | — |
| Epoch Capabilities Index | 136.17 | — |
| LiveBench | — | 46.9% |
Math Qwen3 8B leads
Qwen3 8B: 34.9 (#191), Sonar: 33.7 (#200)
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 56.1% | — |
| LiveBench Math | — | 41.6% |
Knowledge Not comparable
Qwen3 8B: 36.1 (#173), Sonar: —
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| GPQA Diamond | 56.8% | — |
| Vectara Hallucination Rate | 4.8% | — |
Instruction Following Not comparable
Qwen3 8B: —, Sonar: 71.4 (#150)
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| LiveBench Instruction Following | — | 76.2% |
Long Context Not comparable
Qwen3 8B: 37.9 (#210), Sonar: —
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| Fiction.LiveBench | 62.1% | — |
Writing & Preference Not comparable
Qwen3 8B: —, Sonar: 52.6 (#167)
| Benchmark | Qwen3 8B | Sonar |
|---|---|---|
| LiveBench Language | — | 44.1% |
Frequently asked questions
Is Qwen3 8B better than Sonar?
Sonar is the stronger model overall, scoring 38.5 to 33.7 on the Noometry Index. Qwen3 8B costs 3.2× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Which is cheaper, Qwen3 8B or Sonar?
Qwen3 8B is cheaper. It lists at $0.18 per million input tokens and $0.70 per million output tokens; Sonar lists at $1 and $1.
Is Qwen3 8B or Sonar better for coding?
Sonar scores higher on coding benchmarks: 35.7 versus 34.0 in the Noometry coding category.
Which has the bigger context window?
Qwen3 8B does, with 131K tokens against 128K.
How many benchmarks do Qwen3 8B and Sonar share?
0 benchmarks have published results for both models. Qwen3 8B has 11 scored results on Noometry and Sonar has 7.