ModelsCompareBest forBenchmarksStatusPricingAPI
Compare / head-to-head

DeepSeek V3vsDeepSeek V3.2

DeepSeek V3.2 leads 5 of 5 shared benchmarks. 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 – 5
DeepSeek V3.2 leads
Cheaper per token
list price, input + output
Larger context
DeepSeek V3.2
131K tokens
Provider uptime (30d)
100%
DeepSeek
DeepSeek V3
DeepSeek · released 2024-12-26
text
Context
128K
Max out
8K
Input /1M
Output /1M
Cached /1M
Scores
8 · 7 core
DeepSeek V3.2
DeepSeek · released 2025-12-01
text
Context
131K
Max out
66K
Input /1M
Output /1M
Cached /1M
Scores
8 · 7 core
Quality

Benchmark matrix

BenchmarkDeepSeek V3DeepSeek V3.2ΔEdge
Reported by both · 5
SWE-bench Verified42% 73.1% -31.1 ptDeepSeek V3.2
GPQA Diamond59.1% 82.4% -23.3 ptDeepSeek V3.2
MMLU-Pro75.9% 85% -9.1 ptDeepSeek V3.2
AIME 202525% matharena93.1% -68.1 ptDeepSeek V3.2
LMArena Elo1332.4 1424.8 -92.4DeepSeek V3.2
Only DeepSeek V3 reports · 2
MMLU88.5% not reported
OTIS Mock AIME 2024-2025 (Epoch AI run)15.8% epochnot reported
Only DeepSeek V3.2 reports · 2
AIME 2026not reported94.2% matharena
Humanity's Last Exam (no tools)not reported25.1%
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 minus DeepSeek V3.2 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

SpecDeepSeek V3DeepSeek V3.2Edge
Context window128K tokens131K tokensDeepSeek V3.2
Max output8K tokens66K tokensDeepSeek V3.2
Input price / 1M
Output price / 1M
Cached input / 1M
Input + output / 1M
Lower is cheaper. List prices; batch, tool and regional fees excluded.
ModalitiestexttextTie
Released2024-12-262025-12-01
Cited benchmark scores88
Reliability

DeepSeek status

All providers
More matchups

DeepSeek V3 vs …

Built by Respan
Which one wins on your data?

Public benchmarks are a starting point. Run DeepSeek V3 and DeepSeek V3.2 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.