Skip to content

DeepSeek

DeepSeek V4 Pro (2026-04-23)

DeepSeek V4 Pro (2026-04-23) is a model with open weights from DeepSeek. On the Vals AI Legal Research Bench it reaches 23.08 percent all-pass, position 67 of 91, within a standard error of ±2.93 percentage points at $0.68 per task, as at 9 November 2026 [1]. It does not pass the compliance gate: No EU guarantee: Microsoft Azure (Foundry Models): Global Standard, no EU Data Zone Standard.

  • Generally available
  • Open weights
  • 23 April 2026
PrudentScore
-No score: the gate is not open
Compliance gate
RejectedNo EU guarantee: Microsoft Azure (Foundry Models): Global Standard, no EU Data Zone Standard
Cost per task
$0.68Vals AI, per completed task [1]

01

Compliance gate

Why

  • No EU guarantee: Microsoft Azure (Foundry Models): Global Standard, no EU Data Zone Standard
  • DeepSeek own API: personal data is "directly collect, process and stor in People's Republic of China" (privacy policy) and the terms are governed by PRC law; no EU contracting party or EU processing.

The gate only opens when all three requirements are covered: processing in the EU, a contracting party in the EU or EEA, and a contractual exclusion of training on customer data. The gate tests a route, not a model name: the same model can pass through one provider and fail through another. Where a provider publishes no figure, we give no estimate.

02

Routes and EU processing

The gate tests a route (processor, region and contract), not a model name. Per route: processing in the EU, the contracting party and training on customer data.

Routes and EU processing
ProcessorRegion or deploymentEU processingSource
Microsoft Azure (Foundry Models)Global Standard, no EU Data Zone StandardNoSourceRead 10 November 2026

03

Scores per benchmark

Every measurement with its source, measurement date and the day we read it. The position counts only the models measured on the same benchmark and version.

  • Legal Research BenchVersion 1 · Counts towards the PrudentScore
    23.08%± 2.9367 of 91

    Vals AI, Legal Research Bench, v1, 9 November 2026 · Source · Measured 9 November 2026 · Read 10 November 2026 · cost per task $0.68

  • Harvey LAB (Vals implementation)Version 1 · Counts towards the PrudentScore
    3.75%± 1.4341 of 80

    Vals AI, Harvey LAB (Vals implementation), v1, 18 October 2026 · Source · Measured 18 October 2026 · Read 10 November 2026 · cost per task $0.74

  • 80.32%± 0.4780 of 113

    Vals AI, LegalBench (Vals leaderboard), v1, 13 October 2026 · Source · Measured 13 October 2026 · Read 10 November 2026

  • CaseLaw v2Version 2
    59.38%± 1.5027 of 54

    Vals AI, CaseLaw v2, v2, 4 May 2026 · Source · Measured 4 May 2026 · Read 10 November 2026

  • TaxEval v2Version 2
    72.08%± 0.8862 of 108

    Vals AI, TaxEval v2, v2, 1 September 2026 · Source · Measured 1 September 2026 · Read 10 November 2026

  • 25.17%± 3.0644 of 82

    Vals AI, Tax Agent Bench, v1, 9 November 2026 · Source · Measured 9 November 2026 · Read 10 November 2026 · cost per task $0.72

  • CorpFin v2Version 2
    61.38%± 0.9655 of 108

    Vals AI, CorpFin v2, v2, 12 August 2026 · Source · Measured 12 August 2026 · Read 10 November 2026

  • Immigration BenchNot measured
  • Access to JusticeNot measured
  • 62.92%± 1.2626 of 50

    Vals AI, Public Benefits Bench, v1.1, 12 October 2026 · Source · Measured 12 October 2026 · Read 10 November 2026 · cost per task $0.46

  • Vals Legal IndexNot measured
  • Vals IndexVersion 2.1
    38.63%± 1.1643 of 50

    Vals AI, Vals Index, v2.1, 18 October 2026 · Source · Measured 18 October 2026 · Read 10 November 2026 · cost per task $1.31

  • 89.39%± 1.6530 of 109

    Vals AI, GPQA Diamond (Vals-run), v1, 1 September 2026 · Source · Measured 1 September 2026 · Read 10 November 2026

  • 87.25%± 0.3432 of 110

    Vals AI, MMLU Pro (Vals-run), v1, 1 September 2026 · Source · Measured 1 September 2026 · Read 10 November 2026

  • Artificial Analysis Intelligence IndexNot measured

Source: Vals AI (Legal Research Bench, Harvey LAB (Vals implementation), LegalBench (Vals leaderboard), CaseLaw v2, …) · Licence: CC BY 4.0 · Adaptation: Prudai selected the models, translated the explanations, added the gate verdict and the PrudentScore and rounded figures to two decimals.

The PrudentScore is a composition by Prudai B.V. based on publicly published figures and does not represent the view or endorsement of Vals AI, Harvey or Artificial Analysis. Figures from Artificial Analysis appear in running text only, with attribution and a link, as that party's terms require [9].

04

Key facts

Developer
DeepSeek
Model family
DeepSeek
Release date
23 April 2026
Status
Generally available
Context window
1M tokens
Max output tokens
384K tokens
List price
$1.32 input · $3.96 outputper million tokens · Read 10 November 2026
Weights
Open

- = not yet recorded in the register.

05

Frequently asked questions about this model

Does DeepSeek V4 Pro (2026-04-23) pass the compliance gate for legal work?

No. No EU guarantee: Microsoft Azure (Foundry Models): Global Standard, no EU Data Zone Standard DeepSeek own API: personal data is "directly collect, process and stor in People's Republic of China" (privacy policy) and the terms are governed by PRC law; no EU contracting party or EU processing. The gate tests a route, not a model name; the sources are on this page.

What does DeepSeek V4 Pro (2026-04-23) cost per task?

Vals computes $0.68 per completed task on the Legal Research Bench [1]. The list price is $1.32 per million input tokens and $3.96 per million output tokens.

How does DeepSeek V4 Pro (2026-04-23) score on legal research?

23.08 percent all-pass on the Vals AI Legal Research Bench, position 67 of 91 measured models, as at 9 November 2026 [1]. That benchmark measures US law; an independent, published benchmark for Dutch legal work does not exist yet.

Related models

Claude and Anthropic, GPT and OpenAI, Gemini and Google, Grok and xAI, Kimi and Moonshot AI, Muse and Meta, Mistral, DeepSeek, GLM and Z.ai, Harvey, Vals AI and Artificial Analysis are trademarks of their respective owners. They are named for identification only and this implies no partnership, sponsorship or endorsement.

Brand marks appear for identification only and imply no partnership, sponsorship or endorsement.