Perplexity, proprietary
Sonar
Sonar by Perplexity ranks 187th of 354 ranked models on the Noometry Index as of October 2026, with a score of 38.5. Its strongest category is instruction following, where it ranks 150th. API pricing starts at $1 per million input tokens and $1 per million output tokens, with a 128K-token context window.
Last verified
Specifications
- Noometry rank
- #187 of 354
- Index score
- 38.5
- Evidence
- Confirmed 7 results
- Provider
Perplexity- Released
- January 1, 2024
- Weights
- Proprietary
- Reasoning
- No
- Context window
- 128K
- Max output
- 4K
- Input price
- $1 / M
- Output price
- $1 / M
- Blended price
- $1 / M
- Output speed
- Not measured
- Value
- #122 of 219
- Knowledge cutoff
- September 2025
- Input
- text
Category scores
Each category score combines every public result we have in that category.
- Coding 35.7
- Reasoning 21.1
- Math 33.7
- Instruction Following 71.4
- Writing & Preference 52.6
| Category | Score | Rank | Results |
|---|---|---|---|
| Coding | 35.7 | #221 | 1 |
| Reasoning | 21.1 | #227 | 2 |
| Math | 33.7 | #200 | 1 |
| Instruction Following | 71.4 | #150 | 1 |
| Writing & Preference | 52.6 | #167 | 1 |
Strengths and weaknesses
Categories where Sonar places highest and lowest among the models ranked in each, with its score against that category's median.
Strongest categories
| Category | Score | vs median | Rank |
|---|---|---|---|
| Instruction Following | 71.4 | +0.1 | #150 of 305, top 50% |
| Writing & Preference | 52.6 | −1.2 | #167 of 312, top 54% |
Closest competitors
The models ranked just above and below Sonar. When scores are this close, price and speed are often the better way to choose.
| Model | Rank | Score | Blended $/M | Speed | |
|---|---|---|---|---|---|
| Olmo 3 32b Think | #183 | 38.7 | — | — | Compare |
| Hunyuan Large 2025 02 10 | #184 | 38.6 | — | — | Compare |
| Trinity Large Thinking | #185 | 38.6 | $0.39 | — | Compare |
| GPT-5.1-Codex | #186 | 38.6 | $3.44 | — | Compare |
| MiniMax-M2.5 | #188 | 38.3 | $0.52 | 49 | Compare |
| Nova Premier 1.0 | #189 | 38.3 | $5 | 9 | Compare |
| Qwen3-Coder 480B-A35B Instruct | #190 | 38.1 | $3 | 67 | Compare |
| Olmo 3.1 32b Think | #191 | 37.9 | — | — | Compare |
Sponsored placements are available on pages like this one. Advertise on Noometry
Benchmark results
Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.
Coding
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LiveBench Coding | 35.1% | #29 of 39, top 75% | Epoch AI |
Reasoning
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LiveBench Reasoning | 46.3% | #22 of 39, top 57% | Epoch AI | ||
| LiveBench Data Analysis | 37.9% | #32 of 39, top 83% | Epoch AI | ||
| LiveBench | 46.9% | #23 of 39, top 59% | Epoch AI |
Math
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LiveBench Math | 41.6% | #26 of 39, top 67% | Epoch AI |
Instruction Following
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LiveBench Instruction Following | 76.2% | #13 of 39, top 34% | Epoch AI |
Writing & Preference
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LiveBench Language | 44.1% | #17 of 39, top 44% | Epoch AI |
API pricing by provider
| Route | Input $/M | Output $/M | Cached input $/M | Checked |
|---|---|---|---|---|
| openrouter | $1 | $1 | — | 2026-10-10 |
| perplexity | $1 | $1 | — | 2026-10-10 |
Compare Sonar
- Sonar vs GPT-5.1-Codex
- Sonar vs MiniMax-M2.5
- Sonar vs Trinity Large Thinking
- Sonar vs Nova Premier 1.0
- Sonar vs Hunyuan Large 2025 02 10
- Sonar vs Qwen3-Coder 480B-A35B Instruct
- Sonar vs GPT-6 Astra
- Sonar vs Claude Fable 5.1
- Sonar vs Gemini 3.8 Flash
- Sonar vs Kimi K3
- Sonar vs Grok 4.6
- Sonar vs Qwen3.8 Max
- Sonar vs GLM-5.3
- Sonar vs Muse Spark 1.3
Other Perplexity models
Frequently asked questions
How good is Sonar?
Sonar by Perplexity ranks 187th of 354 ranked models on the Noometry Index as of October 2026, with a score of 38.5. Its strongest category is instruction following, where it ranks 150th. API pricing starts at $1 per million input tokens and $1 per million output tokens, with a 128K-token context window.
How much does Sonar cost?
Sonar costs $1 per million input tokens and $1 per million output tokens on Perplexity's own API.
What is Sonar's context window?
Sonar accepts up to 128K tokens of input and can write up to 4K tokens in one response.
Is Sonar open source?
No. Sonar is proprietary and available only through Perplexity's API and partner platforms.
What are Sonar's strengths and weaknesses?
Relative to other ranked models, Sonar places best in instruction following, writing & preference and lowest in coding, reasoning.
What is Sonar best at?
Its best category is instruction following, where it ranks 150th on Noometry.