Skip to content

DeepSeek V4 Pro 0813

ModelOpen sourceActive
by DeepSeek · family “deepseek-v4-pro”

DeepSeek V4 Pro 0813 is a large-scale mixture-of-experts model from DeepSeek. This is the GA release of DeepSeek V4 Pro. (description from the OpenRouter listing)

in: textout: textReasoningTool useStructured output
Epoch Capabilities Index
155.4 #24 of 268
90% CI 153.7 – 157.3
Listing ↗
Released
13 Aug 2026
Context window
1.05M tokens
Max output
944K tokens
Input $ / 1M tokens
$0.39
Output $ / 1M tokens
$3.49
Cached input $ / 1M
$0.32
Each benchmark shown separately with its own source

Benchmark results (11)

All benchmarks
BenchmarkDomainScorevs best recordedSettingRunSource
Mystery Game Puzzlesgames43.0% ±5.0
51%
max19 Aug 2026Epoch ↗
FrontierMath-Tiers-1-3-v2-Privatemath64.6% ±2.8
69%
max19 Aug 2026Eval log ↗
FrontierMath-Tier-4-v2-Privatemath26.8% ±7.0
27%
max19 Aug 2026Eval log ↗
FrontierCodecoding28.5%
53%
high—External ↗
Chess Puzzlesgames47.0% ±5.0
65%
max18 Aug 2026Epoch ↗
SimpleQA Verifiedknowledge52.9% ±1.6
70%
max27 Aug 2026Eval log ↗
ARC-AGI-2reasoning61.3%
64%
max—External ↗
WeirdMLcoding66.2% ±0.0
71%
max—External ↗
OTIS Mock AIME 2024-2025math98.6% ±1.0
99%
max18 Aug 2026Epoch ↗
GPQA diamondscience91.7% ±1.5
96%
max18 Aug 2026Epoch ↗
ARC-AGIreasoning90.0%
91%
max—External ↗

Source: Epoch AI Benchmarking Hub (CC BY 4.0). “External” rows are leaderboard results Epoch collects from third parties. Best reported setting per benchmark is shown.

API list price over time

Price history

$ per 1M tokens

5 recorded prices on 2 Sep 2026 (OpenRouter listing + Internet Archive snapshots); re-read on every data refresh, most recently 39m ago. Steps show when the price changed.

Events