Data points: 96
Model compare
The readout for DeepSeek V3 and DeepSeek V3.2, before the detailed comparison sheet.
Weighted outcome: DeepSeek V3.2. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.
Decision read
DeepSeek V3.2
DeepSeek V3.2 has the higher weighted result; Model A / B score 38.7 to 61.3.
Evidence depth
96 data points
Includes 20 benchmark rows, 0 audit samples, and 6 provider examples.
Selection signal
Start with DeepSeek V3.2
The charts below split 26 high-signal samples across speed, scores, and audit health.
Switch either side of this report to compare another model with the same LMSpeed data pipeline.
Select a different model to open a new comparison URL.
This report only uses LMSpeed data for DeepSeek V3 and DeepSeek V3.2: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.
| Model compare | DeepSeek V3 | DeepSeek V3.2 |
|---|---|---|
| Overall leader | Contender | Leading |
| Weighted overall score | 38.7 pts | 61.3 pts |
| Benchmark category leads | 2 categories | 4 categories |
| Operational advantages | Average speed, First-token latency, Free providers | Cheapest input price, Provider coverage |
| Context window | No data | 163.8K tokens |
| Max output | No data | 65.5K tokens |
| Modalities | No data | Input Text Output Text |
| Features |
The overall result weights benchmark capability categories at 80% and price, API speed/latency, and availability at 20%. Recent test volume does not affect the winner, and missing benchmark categories are excluded.
| Model compare | DeepSeek V3 | DeepSeek V3.2 |
|---|---|---|
| Developer | No data | DeepSeek |
| Released | No data | Dec 2025 |
| Parameters | No data | No data |
| Tokenizer | No data | DeepSeek |
| Knowledge cutoff | No data | No data |
| OpenRouter ID | No data | deepseek/deepseek-v3.2 |
| References | No data | No data |
This report only uses LMSpeed data for DeepSeek V3 and DeepSeek V3.2: pricing, speed aggregates, third-party benchmark scores, and shared provider samples.
DeepSeek V3
DeepSeek V3 is stronger in benchmark categories (Knowledge, Math) and operational dimensions (Average speed, First-token latency, Free providers).
DeepSeek V3.2
DeepSeek V3.2 is stronger in benchmark categories (Agents, Coding, Reasoning, Instruction following) and operational dimensions (Cheapest input price, Provider coverage).
Third-party benchmark profile synced into LMSpeed; only metrics available for both models are shown.
Compare benchmark category scores on a 0-100 scale. Select a category to inspect the gap.
Avg. score
DeepSeek V3
41.5
Avg. score
DeepSeek V3.2
46.3
DeepSeek V3.2 leads by 4.1
DeepSeek V3.2 leads by 13.7
DeepSeek V3.2 leads by 6.5
DeepSeek V3 leads by 1.4
DeepSeek V3 leads by 0.5
No data
No data
DeepSeek V3.2 leads by 6.5
Metric-level scores with benchmark source, rank depth, confidence, error, and evaluation date where available.
DeepSeek V3
$0.890/M
Rank #37/162 · confidence 4
DeepSeek V3.2
winner$0.420/M
Rank #17/162 · confidence 4
DeepSeek V3
$0.493/M
Rank #39/162 · confidence 4
DeepSeek V3.2
winner$0.315/M
Rank #26/162 · confidence 4
DeepSeek V3
$0.360/M
Rank #59/162 · confidence 4
DeepSeek V3.2
winner$0.280/M
Rank #44/162 · confidence 4
DeepSeek V3
75.2%
Rank #85/125 · confidence 4
DeepSeek V3.2
winner86.2%
Rank #15/125 · confidence 4
DeepSeek V3
3.6%
Rank #173/187 · confidence 4
DeepSeek V3.2
winner22.2%
Rank #50/187 · confidence 4
DeepSeek V3
55.7%
Rank #155/188 · confidence 4
DeepSeek V3.2
winner84.0%
Rank #53/188 · confidence 4
DeepSeek V3
29.0
Rank #81/89 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner39.0
Rank #76/89 · confidence 1 · eval date 2025-12-01
DeepSeek V3
35.9%
Rank #78/115 · confidence 4
DeepSeek V3.2
winner86.2%
Rank #5/115 · confidence 4
DeepSeek V3
35.4
Rank #77/96 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner38.7
Rank #65/96 · confidence 1 · eval date 2025-12-01
DeepSeek V3
35.4%
Rank #110/185 · confidence 4
DeepSeek V3.2
winner38.9%
Rank #82/185 · confidence 4
DeepSeek V3
1.7
Rank #44/47 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner22.1
Rank #26/47 · confidence 1 · eval date 2025-12-01
DeepSeek V3
26.3
Rank #50/56 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner40.4
Rank #36/56 · confidence 1 · eval date 2025-12-01
DeepSeek V3
89.4
Rank #14/90 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner93.5
Rank #5/90 · confidence 1 · eval date 2025-12-01
DeepSeek V3
winner25.4
Rank #54/90 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
24.2
Rank #62/90 · confidence 1 · eval date 2025-12-01
DeepSeek V3
14.2
Rank #85/99 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner24.7
Rank #67/99 · confidence 1 · eval date 2025-12-01
DeepSeek V3
3.6
Rank #93/96 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner10.5
Rank #69/96 · confidence 1 · eval date 2025-12-01
DeepSeek V3
55.7
Rank #90/96 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner75.1
Rank #70/96 · confidence 1 · eval date 2025-12-01
DeepSeek V3
1150.0
Rank #53/66 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner1204.0
Rank #43/66 · confidence 1 · eval date 2025-12-01
DeepSeek V3
34.8
Rank #81/86 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner49.0
Rank #58/86 · confidence 1 · eval date 2025-12-01
DeepSeek V3
22.8
Rank #78/84 · confidence 1 · eval date 2024-12-26
DeepSeek V3.2
winner78.9
Rank #56/84 · confidence 1 · eval date 2025-12-01
Latest completed audits from shared providers, with four safety and integrity score groups plus report links.
| Provider | DeepSeek V3 | DeepSeek V3.2 |
|---|---|---|
| No completed audits are available from shared providers yet. | ||
Speed aggregates and input/output pricing share each provider row for real API selection and migration cost checks.
| Provider | DeepSeek V3 | DeepSeek V3.2 |
|---|---|---|
156 tests | DeepSeek V3 speed / latency 15 tok/s / 2662ms input / output No data | DeepSeek V3.2 speed / latency 26 tok/s / 11125ms input / output No data |
130 tests | DeepSeek V3 speed / latency 36 tok/s / 1849ms input / output No data | DeepSeek V3.2 speed / latency 27 tok/s / 708ms input / output No data |
75 tests | DeepSeek V3 speed / latency 41 tok/s / 1245ms input / output No data | DeepSeek V3.2 speed / latency 27 tok/s / 1295ms input / output No data |
45 tests | DeepSeek V3 speed / latency N/A / N/A input / output No data | DeepSeek V3.2 speed / latency 235 tok/s / 1787ms input / output No data |
40 tests | DeepSeek V3 speed / latency N/A / N/A input / output No data | DeepSeek V3.2 speed / latency 29 tok/s / 1828ms input / output No data |
DeepSeek V3 deepseek-v3-all speed / latency No data input / output $0/request | DeepSeek V3.2 deepseek-v3.2-speciale speed / latency No data input / output $0/request |
Weighted outcome: DeepSeek V3.2. Benchmark capability categories carry 80%, while price, API performance, and availability carry 20%.
| None listed |
Text inputText outputTool callingStructured outputsJSON modeReasoning |
