Model

DeepSeek-V4-Pro (Think High)

OPEN

DeepSeek · DeepSeek-V4 family · released Apr 24, 2026

Fast/low-latency mode of the DeepSeek-V4-Pro preview checkpoint, the default reasoning tier a user gets without selecting anything. (middle reasoning-effort tier, called 'Think High' on the official HF model-card benchmark table). DeepSeek repriced deepseek-v4-pro (now serving the V4-Pro-0813 GA checkpoint) to $1.32/$3.96 per MTok standard, $0.66/$1.98 off-peak, effective Aug 16, 2026.

ReasoningCodingVisionFunction callingTool useAgentic

2308.5

Elo · rank #60

Parameters

1600B

Active params

49B (MoE)

Context

1M tokens

Architecture

Mixture-of-Experts (1.6T total / 49B active), hybrid Compressed Sparse Attention + Heavily Compressed Attention, Manifold-Constrained Hyper-Connections, Muon optimizer

License

MIT

Languages

—

API price (in/out)

$1.32 / $3.96

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

Agents' Last ExamAgents27.6%#19

best: GPT-6 Astra · 59.3%

BrowseCompAgents80.4%#19

best: Kimi K3 · 91.2%

CodeforcesCoding2919#4

best: DeepSeek-V4.1-Flash · 3471

GPQA DiamondReasoning89.1%#39

best: GPT-6 Astra · 96.0%

Humanity's Last ExamReasoning34.5%#47

best: Claude Opus 5.5 · 67.7%

LiveCodeBenchCoding89.8%#9

best: DeepSeek-V4-Pro (Think Max) · 93.5%

MMLU-ProKnowledge87.1%#13

best: Claude Fable 5 · 91.5%

SWE-bench ProCoding54.4%#38

best: Claude Opus 5.5 · 89.9%

SWE-bench VerifiedCoding79.4%#30

best: Claude Opus 5 · 97.0%

Terminal-Bench 2.0Coding63.3%#44

best: DeepSeek-V4.1-Flash · 90.6%

Run it locally

VRAM @ Q4

—

VRAM @ FP16

—

Fits on (Q4)

Multi-node cluster required

Throughput data unavailable.

Quantizations

—

Fine-tune it

Permissive

QLoRA1088.0 GB8× H200 141GB

LoRA3408.0 GBbeyond 8× B200

Full fine-tune25680.0 GBbeyond 8× B200

QLoRA SFT on ~10k samples ≈ $25.34 (8× H200 141GB)

Plan a fine-tune →

DeepSeek-V4 family

Elo progression across releases

DeepSeek-V4-Flash (Non-Think)Apr 20261561.9 DeepSeek-V4-Flash (Think High)Apr 20262165.3 DeepSeek-V4-Flash (Think Max)Apr 20262250.1 DeepSeek-V4-Pro (Non-Think)Apr 20261608.1 DeepSeek-V4-Pro (Think High)Apr 20262308.5 DeepSeek-V4-Pro (Think Max)Apr 20262432.8 DeepSeek-V4-Flash-0731Jul 20262600.5 DeepSeek-V4.1-FlashSep 20262754.1

Resources

HFdeepseek-ai/DeepSeek-V4-Pro DOCShttps://api-docs.deepseek.com/news/news260424/APIAvailable via DeepSeek API

API price $1.32/$3.96 · each benchmark row carries its own source badge (see methodology)