# DeepSeek V4 Flash vs Sonar

> DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 37.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-flash-vs-sonar
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Summary

- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 20.3.
- DeepSeek V4 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $1 / $1 for Sonar.
- DeepSeek V4 Flash accepts more context: 1M tokens versus 128K.
- DeepSeek V4 Flash has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek V4 Flash | Sonar |
|---|---|---|
| Provider | DeepSeek | Perplexity |
| Noometry Index | 53.6 | 37.8 |
| Rank | 32 | 184 |
| Context | 1M | 128K |
| Input $/M | $0.15 | $1 |
| Output $/M | $0.60 | $1 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek V4 Flash: 47.6 (#56)
- Sonar: 35.1 (#217)

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| FrontierCode | 18.8% | — |
| LMArena WebDev | 1581 | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
| LiveBench Coding | — | 35.1% |
| LMArena Coding | 1457 | — |
| ALE-Bench | 1,306 | — |

## Reasoning

- DeepSeek V4 Flash: 53.7 (#25)
- Sonar: 20.3 (#221)

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | 89% | — |
| CritPt | 16.6% | — |
| Chess Puzzles | 33% | — |
| LiveBench Reasoning | — | 46.3% |
| LMArena Hard Prompts | 1443 | — |
| Mystery Game Puzzles | 34% | — |
| DTBench | 90.9% | — |
| LiveBench Data Analysis | — | 37.9% |
| LMCA | 41.7% | — |
| Epoch Capabilities Index | 154.51 | — |
| LiveBench | — | 46.9% |

## Math

- DeepSeek V4 Flash: 60.0 (#31)
- Sonar: 31.3 (#197)

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| OTIS Mock AIME 2024-2025 | 94.4% | — |
| ProofBench | 56% | — |
| LiveBench Math | — | 41.6% |
| LMArena Math | 1427 | — |

## Knowledge

- DeepSeek V4 Flash: 55.5 (#43)
- Sonar: —

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| GPQA Diamond | 91% | — |
| SimpleQA Verified | 33.6% | — |
| LMArena Expert | 1441 | — |

## Multilingual

- DeepSeek V4 Flash: 53.2 (#71)
- Sonar: —

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| LMArena Non-English | 1420 | — |
| LMArena Chinese | 1467 | — |
| LMArena French | 1438 | — |
| LMArena German | 1418 | — |
| LMArena Japanese | 1403 | — |
| LMArena Korean | 1383 | — |
| LMArena Russian | 1428 | — |
| LMArena Spanish | 1437 | — |

## Instruction Following

- DeepSeek V4 Flash: 75.1 (#76)
- Sonar: 71.5 (#146)

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| LiveBench Instruction Following | — | 76.2% |
| LMArena Instruction Following | 1421 | — |

## Long Context

- DeepSeek V4 Flash: 43.8 (#81)
- Sonar: —

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| LMArena Longer Query | 1435 | — |

## Writing & Preference

- DeepSeek V4 Flash: 64.0 (#59)
- Sonar: 52.6 (#166)

| Benchmark | DeepSeek V4 Flash | Sonar |
|---|---|---|
| LMArena Text | 1432 | — |
| LMArena Creative Writing | 1403 | — |
| EQ-Bench Creative Writing | 1559 | — |
| LMArena Multi-Turn | 1449 | — |
| LiveBench Language | — | 44.1% |

## FAQ

### Is DeepSeek V4 Flash better than Sonar?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 37.8 on the Noometry Index.

### Which is cheaper, DeepSeek V4 Flash or Sonar?

DeepSeek V4 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Sonar lists at $1 and $1.

### Is DeepSeek V4 Flash or Sonar better for coding?

DeepSeek V4 Flash scores higher on coding benchmarks: 47.6 versus 35.1 in the Noometry coding category.

### Which has the bigger context window?

DeepSeek V4 Flash does, with 1M tokens against 128K.

### How many benchmarks do DeepSeek V4 Flash and Sonar share?

0 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Sonar has 7.