ModelsCompareBest forBenchmarksStatusPricingAPI
Compare / head-to-head

DeepSeek V3.2vsGPT-4.1

DeepSeek V3.2 leads 2 of 2 shared benchmarks. GPT-4.1 has the larger context window (1.0M tokens).

Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.

Shared benchmarks
2 – 0
DeepSeek V3.2 leads
Cheaper per token
list price, input + output
Larger context
GPT-4.1
1.0M tokens
Provider uptime (30d)
100% · 95.0311%
DeepSeek · OpenAI
DeepSeek V3.2
DeepSeek · released 2025-12-01
text
Context
131K
Max out
66K
Input /1M
Output /1M
Cached /1M
Scores
8 · 7 core
GPT-4.1
OpenAI · released 2025-04-14
textvision
Context
1.0M
Max out
33K
Input /1M
$2
Output /1M
$8
Cached /1M
$0.5
Scores
10 · 8 core
Quality

Benchmark matrix

BenchmarkDeepSeek V3.2GPT-4.1ΔEdge
Reported by both · 2
SWE-bench Verified73.1% 54.6% +18.5 ptDeepSeek V3.2
GPQA Diamond82.4% 66.3% +16.1 ptDeepSeek V3.2
Only DeepSeek V3.2 reports · 5
MMLU-Pro85% not reported
AIME 202593.1% not reported
AIME 202694.2% matharenanot reported
Humanity's Last Exam (no tools)25.1% not reported
LMArena Elo1424.8 not reported
Only GPT-4.1 reports · 7
MMLUnot reported90.2%
AIME 2024not reported48.1%
FrontierMath Tiers 1-3 v2 (Epoch AI run)not reported6% epoch
OTIS Mock AIME 2024-2025 (Epoch AI run)not reported38.3% epoch
SimpleQA Verifiednot reported31.1% epoch
SWE-bench Verified (Epoch AI run)not reported48.5% epoch
Video-MME (long, no subtitles)not reported72%
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 GPT-4.1 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 V3.2GPT-4.1Edge
Context window131K tokens1.0M tokensGPT-4.1
Max output66K tokens33K tokensDeepSeek V3.2
Input price / 1M$2
Output price / 1M$8
Cached input / 1M$0.5
Input + output / 1M
Lower is cheaper. List prices; batch, tool and regional fees excluded.
$10
Modalitiestexttext · visionGPT-4.1
Released2025-12-012025-04-14
Cited benchmark scores810
Reliability

Provider status

All providers
Built by Respan
Which one wins on your data?

Public benchmarks are a starting point. Run DeepSeek V3.2 and GPT-4.1 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.