Model comparison
Qwen3.5-9B vs Sonar
Sonar is the stronger model overall, scoring 38.5 to 33.8 on the Noometry Index. Qwen3.5-9B costs 8.9× 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
- Qwen3.5-9B is cheaper at $0.10 / $0.15 per million input/output tokens, against $1 / $1 for Sonar.
- Qwen3.5-9B accepts more context: 262K tokens versus 128K.
- Qwen3.5-9B has downloadable open weights; the other is API-only.
Side by side
| Qwen3.5-9B | Sonar | |
|---|---|---|
| Provider | Alibaba (Qwen) | Perplexity |
| Noometry Index | 33.8 | 38.5 |
| Released | 2026-02-23 | 2024-01-01 |
| Weights | Open | Proprietary |
| Context window | 262K | 128K |
| Max output | 66K | 4K |
| Input $ / M tokens | $0.10 | $1 |
| Output $ / M tokens | $0.15 | $1 |
| Results tracked | 10 | 7 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Qwen3.5-9B: 35.9 (#217), Sonar: 35.7 (#221)
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| SciCode | 27.5% | — |
| LiveBench Coding | — | 35.1% |
Agentic & Tool Use Not comparable
Qwen3.5-9B: 14.5 (#151), Sonar: —
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| Terminal-Bench | 9.2% | — |
Reasoning Qwen3.5-9B leads
Qwen3.5-9B: 23.1 (#182), Sonar: 21.1 (#227)
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| CritPt | 0.3% | — |
| Chess Puzzles | 12% | — |
| LiveBench Reasoning | — | 46.3% |
| DTBench | 71.2% | — |
| LiveBench Data Analysis | — | 37.9% |
| LMCA | 24.5% | — |
| Epoch Capabilities Index | 139.46 | — |
| LiveBench | — | 46.9% |
Math Qwen3.5-9B leads
Qwen3.5-9B: 34.8 (#192), Sonar: 33.7 (#200)
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| MathArena Final-Answer Competitions | 48.5% | — |
| OTIS Mock AIME 2024-2025 | 61.7% | — |
| LiveBench Math | — | 41.6% |
Knowledge Not comparable
Qwen3.5-9B: 46.0 (#84), Sonar: —
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| GPQA Diamond | 79% | — |
Instruction Following Not comparable
Qwen3.5-9B: —, Sonar: 71.4 (#150)
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| LiveBench Instruction Following | — | 76.2% |
Writing & Preference Not comparable
Qwen3.5-9B: —, Sonar: 52.6 (#167)
| Benchmark | Qwen3.5-9B | Sonar |
|---|---|---|
| LiveBench Language | — | 44.1% |
Frequently asked questions
Is Qwen3.5-9B better than Sonar?
Sonar is the stronger model overall, scoring 38.5 to 33.8 on the Noometry Index. Qwen3.5-9B costs 8.9× less per token, which makes it the better buy when Sonar's lead doesn't matter for your workload.
Which is cheaper, Qwen3.5-9B or Sonar?
Qwen3.5-9B is cheaper. It lists at $0.10 per million input tokens and $0.15 per million output tokens; Sonar lists at $1 and $1.
Is Qwen3.5-9B or Sonar better for coding?
They score almost the same on coding (35.9 vs 35.7); test both on your own repository before choosing.
Which has the bigger context window?
Qwen3.5-9B does, with 262K tokens against 128K.
How many benchmarks do Qwen3.5-9B and Sonar share?
0 benchmarks have published results for both models. Qwen3.5-9B has 10 scored results on Noometry and Sonar has 7.