Compare / head-to-head
DeepSeek V3.2vsSonar Deep Research
DeepSeek V3.2 and Sonar Deep Research do not yet share a public benchmark, so quality cannot be compared apples-to-apples. DeepSeek V3.2 has the larger context window (131K tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0
no overlap yet
Cheaper per token
—
list price, input + output
Larger context
DeepSeek V3.2
131K tokens
Provider uptime (30d)
100% · 100%
DeepSeek · Perplexity
DeepSeek V3.2
DeepSeek · released 2025-12-01
text
- Context
- 131K
- Max out
- 66K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 8 · 7 core
Sonar Deep Research
Perplexity · released 2025-02-14
text
Quality
Benchmark matrix
0 shared · 7 only DeepSeek V3.2 · 6 only Sonar Deep Research| Benchmark | DeepSeek V3.2 | Sonar Deep Research | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 0 | ||||
| No public benchmark is reported for both models yet. | ||||
| Only DeepSeek V3.2 reports · 7 | ||||
| SWE-bench Verified | 73.1% ↗ | not reported | — | — |
| GPQA Diamond | 82.4% ↗ | not reported | — | — |
| MMLU-Pro | 85% ↗ | not reported | — | — |
| AIME 2025 | 93.1% ↗ | not reported | — | — |
| AIME 2026 | 94.2% ↗matharena | not reported | — | — |
| Humanity's Last Exam (no tools) | 25.1% ↗ | not reported | — | — |
| LMArena Elo | 1424.8 ↗ | not reported | — | — |
| Only Sonar Deep Research reports · 6 | ||||
| DRACO Breadth & Depth (Opus 4.6 base) | not reported | 73.1% ↗ | — | — |
| DRACO Citation Quality (Opus 4.6 base) | not reported | 64.6% ↗ | — | — |
| DRACO Factual Accuracy (Opus 4.6 base) | not reported | 67.9% ↗ | — | — |
| DRACO Normalized Score (Opus 4.6 base) | not reported | 70.5% ↗ | — | — |
| DRACO Pass Rate (Opus 4.6 base) | not reported | 72.8% ↗ | — | — |
| DRACO Presentation Quality (Opus 4.6 base) | not reported | 90.3% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is DeepSeek V3.2 minus Sonar Deep Research in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | DeepSeek V3.2 | Sonar Deep Research | Edge |
|---|---|---|---|
| Context window | 131K tokens | 128K tokens | DeepSeek V3.2 |
| Max output | 66K tokens | — | — |
| Input price / 1M | — | $2 ↗ | — |
| Output price / 1M | — | $8 ↗ | — |
| Cached input / 1M | — | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | — | $10 | — |
| Modalities | text | text | Tie |
| Released | 2025-12-01 | 2025-02-14 | — |
| Cited benchmark scores | 8 | 6 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
DeepSeek V3.2 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run DeepSeek V3.2 and Sonar Deep Research on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.