Compare / head-to-head
DeepSeek V4 ProvsGPT-5.2
DeepSeek V4 Pro and GPT-5.2 split their 8 shared benchmarks evenly. DeepSeek V4 Pro is 12.1x cheaper per token. DeepSeek V4 Pro has the larger context window (1M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
4 – 4
even split
Cheaper per token
DeepSeek V4 Pro
12.1x cheaper, input + output
Larger context
DeepSeek V4 Pro
1M tokens
Provider uptime (30d)
100% · 95.4802%
DeepSeek · OpenAI
DeepSeek V4 Pro
DeepSeek · released 2026-08-13
text
Quality
Benchmark matrix
8 shared · 3 only DeepSeek V4 Pro · 6 only GPT-5.2| Benchmark | DeepSeek V4 Pro | GPT-5.2 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 8 | ||||
| GPQA Diamond | 90.9% ↗epoch | 92.4% ↗ | -1.5 pt | GPT-5.2 |
| AIME 2026 | 96.7% ↗matharena ⚠ | 98.3% ↗matharena | -1.6 pt | GPT-5.2 |
| LMArena Elo | 1450.6 ↗ | 1412.4 ↗ | +38.2 | DeepSeek V4 Pro |
| FrontierMath Tier 4 v2 (Epoch AI run) | 2.4% ↗epoch | 31.7% ↗epoch | -29.3 pt | GPT-5.2 |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 45.3% ↗epoch | 67.4% ↗epoch | -22.1 pt | GPT-5.2 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 96.7% ↗epoch | 96.1% ↗epoch | +0.6 pt | DeepSeek V4 Pro |
| SimpleQA Verified | 47% ↗epoch | 37.1% ↗epoch | +9.9 pt | DeepSeek V4 Pro |
| SWE-bench Verified (Epoch AI run) | 77.6% ↗epoch | 73.8% ↗epoch | +3.8 pt | DeepSeek V4 Pro |
| Only DeepSeek V4 Pro reports · 3 | ||||
| DeepSWE v1.1 | 62.7% ↗ | not reported | — | — |
| Humanity's Last Exam (with tools) | 60% ↗ | not reported | — | — |
| Terminal-Bench 2.1 | 87.9% ↗ | not reported | — | — |
| Only GPT-5.2 reports · 6 | ||||
| SWE-bench Verified | not reported | 80% ↗ | — | — |
| AIME 2025 | not reported | 100% ↗ | — | — |
| ARC-AGI-1 (Verified) | not reported | 86.2% ↗ | — | — |
| ARC-AGI-2 (Verified) | not reported | 52.9% ↗ | — | — |
| GDPval (wins or ties) | not reported | 70.9% ↗ | — | — |
| SWE-Bench Pro | not reported | 55.6% ↗ | — | — |
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 V4 Pro minus GPT-5.2 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
Official pricing pages and model cards| Spec | DeepSeek V4 Pro | GPT-5.2 | Edge |
|---|---|---|---|
| Context window | 1M tokens | 400K tokens | DeepSeek V4 Pro |
| Max output | 384K tokens | 128K tokens | DeepSeek V4 Pro |
| Input price / 1M | $0.435 ↗ | $1.75 ↗ | DeepSeek V4 Pro |
| Output price / 1M | $0.87 ↗ | $14 ↗ | DeepSeek V4 Pro |
| Cached input / 1M | $0.004 ↗ | $0.175 ↗ | DeepSeek V4 Pro |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $1.31 | $15.75 | DeepSeek V4 Pro |
| Modalities | text | text · vision | GPT-5.2 |
| Released | 2026-08-13 | 2025-12-11 | — |
| Cited benchmark scores | 11 | 16 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
DeepSeek V4 Pro vs …
Models sharing the most benchmarksMore matchups
GPT-5.2 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run DeepSeek V4 Pro and GPT-5.2 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.