AI Model Profile

DeepSeek V4 Pro

DeepSeek V4 Pro is listed in DeepSeek API docs as a current model supporting thinking and non-thinking modes, JSON output, tool calls, and a 1M context length.

Family
DeepSeek V4
Release date
Unknown
Status
Current
Context window
1M tokens
Output limit
384K tokens maximum
API
yes
Open weights
no
Local/self-hosted
no
Pricing
DeepSeek API pricing checked September 21, 2026 for deepseek-v4-pro (DeepSeek-V4-Pro-0813). Per million tokens, off-peak cache-miss input / cache-hit input / output: $0.66 / $0.022 / $1.98. Peak: $1.32 / $0.044 / $3.96. These are the current API documentation’s displayed dollar rates; peak hours and the unresolved earlier routing announcement are explained below.
Evidence state
Recheck due

Benchmark Caveat

Provider documentation and provider-published capability notes are directional, not universal rankings. Real results depend on prompts, tools, latency targets, pricing tier, safety filters, context length, and workload mix.

Best for

Candidate for DeepSeek API workflows that need the Pro variant and long-context reasoning or coding tests.

Skip if

Skip if your team needs independently benchmarked performance claims or contractual guarantees not present in the official provider documentation.

Strengths

Official docs list a long context window, tool calls, JSON output, and current V4 Pro access.

Weaknesses

This profile does not independently validate benchmark quality, reliability, or latency.

Agent suitability

Candidate for tool-using agents based on official API feature support.

Kingy AI take

Use this profile as a source-backed reference point, not a ranking. Re-check official provider docs before making production or budget decisions.

Verification & Sources

Evidence state
Recheck due
Source links
7
Freshness
Needs recheck: checked July 28, 2026
Last updated
July 28, 2026
What this evidence state means
Definition
The claim was previously checked, but its review window expired or a material change may have invalidated it.
Required provenance
The prior evidence and check date are retained, together with the expiry or change signal that triggered recheck.
Owner
Kingy freshness queue owner and assigned editorial reviewer
Freshness rule
This is already outside its freshness rule. It must not be presented as current until reviewed against current evidence.
Disputes and corrections
Use “Suggest a correction” on the record. Kingy editorial reviews the cited evidence, records material corrections, and changes or removes the state when it is not supported.
Suggest a correction

Form submissions, correction notes, score details, URLs, and analytics events may be stored for editorial review, spam prevention, product improvement, and follow-up. Do not submit secrets, unreleased financials, private customer data, or regulated personal data through these forms.

Full Model Notes

DeepSeek V4 Pro is listed in DeepSeek API docs as a current model supporting thinking and non-thinking modes, JSON output, tool calls, and a 1M context length.

Coding notes

Evaluate coding workflows directly, especially for agentic code assistants.

Reasoning notes

Official docs describe thinking and non-thinking mode support.

Creative notes

Not primarily a creative/media model in this profile.

Research notes

Candidate for long-context research tests with current API docs.

API pricing notes

Pricing rechecked September 21, 2026. Peak hours: 01:00–04:00 and 06:00–10:00 UTC, Monday–Friday, excluding Chinese public holidays. All other hours, including weekends and those holidays, are off-peak. The current pricing table lists V4 Pro separately; the Anthropic-compatible API guide also bills its claude-opus mapping at V4 Pro prices. The September 10 announcement instead said Pro requests would route to V4.1-Flash from September 14. Kingy uses the current pricing table for this dated price observation and retains that source conflict. Actual request routing and account billing were not tested. No paid API call was made. Confirm the returned model and charged usage before relying on a production budget. Earlier profile notes, retained from July 28 and not rechecked in this pricing review: Record returned model, system_fingerprint and usage. Thinking is the default. Publication requires clear AI-output disclosure; DeepSeek privacy materials describe processing and storage in the People's Republic of China.

License notes

Review the official provider terms and model documentation before relying on license or redistribution assumptions.