Crosshair
Leaderboard
Language ModelsOpen weights

DeepSeek V4-Pro

Open-weights MoE flagship (1.6T total / 49B active) with built-in reasoning modes (MIT).

textcodeOfficial site
Crosshair Index
70.7
#9 of 32 · Language Models
Provider
DeepSeek
Released
2026-04-24
Parameters
1600B
Context
1M tokens

Token pricing

Input
$0.435 /1M
Output
$0.87 /1M
Cache read
$0.004 /1M

USD per 1M tokens · cache read = cached input (hit), cache write = caching surcharge · official list pricing (June 2026).

Industry skill web

Professional-domain strengths, composed from the benchmarks relevant to each field. Highlight an axis to see the benchmarks behind it.

SoftwareIB / FinanceLawMedicineResearchConsultingAccounting

Software Engineering

87
skill

Shipping working code against real repositories: bug fixes, feature patches, and competitive programming under tests.

SWE-bench Verified80.6%
LiveCodeBench93.5%

Scorecard

BenchmarkScoreSourceStatus
GPQA Diamond
Reasoning
90.1%DeepSeek — V4-Pro model cardvendorunverified
Humanity's Last Exam
Frontier
37.7%DeepSeek — V4-Pro model cardvendorunverified
SWE-bench Verified
Agentic Coding
80.6%DeepSeek — V4-Pro model cardvendorunverified
LiveCodeBench
Coding
93.5%bestDeepSeek — V4-Pro model cardvendorunverified
MMLU-Pro
Knowledge
87.5%bestDeepSeek — V4-Pro model cardvendorunverified
LMArena Elo
Human Preference
1,457LMArena (arena.ai)3rd-partyunverified
AA Intelligence Index
Composite
52Artificial Analysis3rd-partyunverified
Corporate Finance
Finance
61.4%Vals AI — CorpFin v23rd-partyunverified
LegalBench
Law
80.3%Vals AI — LegalBench3rd-partyunverified
TaxEval
Tax & Accounting
72.1%Vals AI — TaxEval v23rd-partyunverified
Medical Coding
Medicine
40.5%Vals AI — MedCode3rd-partyunverified