Back to the index
DeepSeek · DeepSeek
No. 24

DeepSeek-V4 Pro

Open MIT flagship — 1.6T-parameter MoE with native vision and a 1M-token context at throwaway prices.

OpenOpen weightsText, Image
Field notes
Context window
1Mtokens
Max output
384Ktokens
Input price
$0.43/ 1M tokens
Output price
$0.87/ 1M tokens
Cached input
$0.04/ 1M tokens
Modalities
Text, Image
Knowledge cutoff
Jan 2026
Released
Apr 2026
Parameters
1600B
License
MIT
Approx. speed
Overview

What this model is

DeepSeek-V4 Pro is DeepSeek's open-weight flagship (1.6T-parameter MoE, 49B active, MIT) and the headline of the V4 generation. It expands the context window 8x to 1M tokens, adds native image input, and introduces a refined DeepSeek Sparse Attention (a CSA/HCA hybrid) that cuts long-context compute and KV-cache footprint dramatically. DeepSeek advertises a chart-topping 80.6% on SWE-bench Verified, but independent contamination-resistant runs land far lower (~58%) and it trails Claude Opus on SWE-bench Pro, so it is a spectacular-value workhorse rather than a frontier-reliable autonomous coder. Its API speaks both OpenAI and Anthropic formats, dropping into Claude Code without a proxy.

Strengths

  • Among the lowest frontier-class prices in the industry
  • 1M-token context via efficient DeepSeek Sparse Attention
  • Native image input (new to the V4 generation)
  • Very cheap cached input; MIT open weights
  • Strong math and general reasoning value

Trade-offs

  • Vendor benchmarks overstate real-world coding reliability
  • Agentic-coding reliability below the Claude/GPT frontier
  • Vision quality trails dedicated multimodal flagships
  • Large 1.6T MoE — serving cost and latency

Best for

  • High-volume, cost-sensitive workloads at scale
  • Long-context document and repo processing
  • Math and reasoning tasks
  • Budget Claude Code-compatible backend
  • Self-hosted / on-prem deployments

Not ideal for

  • Unsupervised long-horizon autonomous coding
  • Mission-critical frontier reasoning
Capabilities

Capability profile

Normalized 0–100 scores, comparable across the whole catalog.

Reasoning85Coding73Math91Writing81Knowledge85Speed60Agentic69Vision66Multilingual83Long Context87
Reasoning
85
Coding
73
Math
91
Writing
81
Knowledge
85
Speed
60
Agentic
69
Vision
66
Multilingual
83
Long Context
87
Benchmarks

How it scores

Public benchmark results, with independent third-party results where available. Bars normalize percentages to 100 and Elo ratings to a 1500 ceiling.

Independent index
Artificial Analysis Intelligence Index
Composite of ~9–10 independent evals · Artificial Analysis
54/ 100
BenchmarkResult
Artificial Analysis Intelligence Indexindependent
GeneralArtificial Analysis
54%
SWE-bench Verifiedindependent
CodingSWE-bench
80.6%
SWE-bench Proindependent
CodingSWE-bench Pro
55.4%
GPQA Diamondindependent
ReasoningIndependent aggregators
83%
AIME 2025
MathVendor-reported
93%
LiveCodeBenchindependent
CodingLiveCodeBench
78%
MMLU-Pro
KnowledgeVendor-reported
86%
DeepSWEindependent
CodingDatacurve
8%
LMArena Eloindependent
GeneralLMArena
1462Elo
Alternatives

Comparable models

Models in a roughly similar class — worth weighing against this one.