VerdictPal
Join

Models / DeepSeek

DeepSeek V4 Pro (Non-reasoning)

DeepSeek V4 · Released 2026-04-24

AA Intelligence Index 31.2. Listed on Artificial Analysis with API pricing $0.435/$0.87 per 1M tokens. Benchmark rows sync from the AA snapshot — see the ledger on this page.

Frontier model tracked on Artificial Analysis. Use the benchmark ledger below for sourced scores; we do not infer numbers beyond published rows.

#92 of 131 on AA Index · current snapshot
31 AA Index · 10 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite46
ContextUnlisted
Input / 1M$0.43
Output / 1M$0.87
Knowledge cutoffunpublished
Output speed71 t/s
TTFT (default API)1.00s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
72
Coding
42
Agentic
91
Long context
45

63average across 4 tested families

Benchmark placements

Where DeepSeek V4 Pro (Non-reasoning) places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
τ³-Bench BankingAgentic#20/ 14391Artificial Analysis2026-07-28
AA output speedPerformance#50/ 149Artificial Analysis2026-07-28
AA time to first tokenPerformance#51/ 149Artificial Analysis2026-07-28
SciCodeCoding#83/ 14442Artificial Analysis2026-07-28
IFBenchReasoning#101/ 12146Artificial Analysis2026-07-28
Terminal-BenchAgentic#103/ 14836Artificial Analysis2026-07-28
Artificial Analysis Intelligence IndexReasoning#106/ 14731Artificial Analysis2026-07-28
AA-LCRLong context#121/ 14345Artificial Analysis2026-07-28
Humanity's Last ExamReasoning#125/ 1448Artificial Analysis2026-07-28
GPQA DiamondReasoning#131/ 14972Artificial Analysis2026-07-28
Family context

The three highest-scoring carded models in each capability family. Where DeepSeek V4 Pro (Non-reasoning) shows up, it's highlighted.

ReasoningDeepSeek V4 Pro (Non-reasoning) · 72
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingDeepSeek V4 Pro (Non-reasoning) · 42
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticDeepSeek V4 Pro (Non-reasoning) · 91
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextDeepSeek V4 Pro (Non-reasoning) · 45
  1. 1
    GPT-5 (high)
    76
  2. 2
    GPT-5.2 Codex (xhigh)
    76
  3. 3
    Kimi K3
    75
Newest receipts

Benchmark rows added to the public ledger for DeepSeek V4 Pro (Non-reasoning) in the last 120 days. Older rows live in the full table below.

10 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context45%Artificial Analysis4
AA output speedPerformance71 t/sArtificial Analysis4
AA time to first tokenPerformance1.00sArtificial Analysis4
Artificial Analysis Intelligence IndexReasoning31.2%Artificial Analysis4
GPQA DiamondReasoning71.7%Artificial Analysis4
Humanity's Last ExamReasoning7.7%Artificial Analysis4
IFBenchReasoning45.78%Artificial Analysis4
SciCodeCoding42.4%Artificial Analysis4
τ³-Bench BankingAgentic91.23%Artificial Analysis4
Terminal-BenchAgentic36.36%Artificial Analysis4
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning72
Coding42
Agentic91
Long context45
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotxAIMiniMaxMetaZhipuXiaomiOtherMeituanCohereNVIDIA

Full benchmark ledger

Every catalog benchmark for DeepSeek V4 Pro (Non-reasoning). Scores link to the original source; gaps mean no public row exists yet.

10 sourced rows

10 of 32 catalog benchmarks have a sourced row for DeepSeek V4 Pro (Non-reasoning).

Reasoning

Coding

Agentic

Long context

Performance

Sources