Scheduled for retirement on Feb 20, 2028. Still available at the prices below until then.

DeepSeek V4 Pro 0423

DeepSeekReleased Apr 24, 2026

Tool use Reasoning Prompt caching Structured output

Compare DeepSeek V4 Pro 0423 Count tokens for DeepSeek V4 Pro 0423

Context

Max output

Serving providers

Cheapest input

$0.435 /1M

On-demand listed rates, OpenRouter routed quotes and Vercel reseller quotes. Enterprise commitments (committed-use discounts, provisioned throughput, negotiated rates) can be materially lower.

Benchmarks

independent evaluations · source: Artificial Analysis and Hugging Face leaderboards

Speed & latency

Output speed

100 t/s

tokens / second

Time to first token

1.64s

latency

Headline indices

Intelligence Index

30.4

of 100

Individual benchmarks · click a row for effort variants

Benchmark

Best score

Effort

Cost / task

Aa Analyst Agent

0.2

default

—

AA-BriefcaseUndated

22%

default

—

APEX Agents

24%

default

—

Automation Bench

0.6

default

—

CritPt

13%

default

—

GDPval (normalized)Frozen Aug 18, 2026

33%

default

—

GPQA Diamond

91%

high

—

Harvey Lab

0.8

default

—

Humanity's Last Exam

38%

default

—

IFBench

76%

default

—

IT-Bench SRE

38%

default

—

Long-Context Reasoning

75%

default

—

Omniscience

-10.6

high

—

Omniscience Accuracy

0.4

default

—

Omniscience Non Hallucination

0.1

reasoning: false

—

SciCode

51%

default

—

SWE-bench Pro

55%

default

—

SWE-bench Verified

81%

default

—

Terminal Bench40

0.1

default

—

TerminalBench Hard

46%

default

—

TerminalBench v2.1

65%

high

—

τ-bench Banking

30%

default

—

τ²-bench

96%

default

—

We cannot tell whether this benchmark is still being published. Treat these scores as undated.

Artificial Analysis last published scores for this benchmark on Aug 18, 2026. Scores here are frozen at that date, and models released since are not on it.

Pricing detail

cost beyond the standard rate · source: models.dev

Reasoning tokens

$1.98

per 1M · thinking output

Pricing by serving provider

standard tier · per 1M tokens · click a provider to expand

Input pricing runs from $0.435 to $2.40 per 1M tokens across 16 providers, so the dearest route costs 452% more than the cheapest for the same model.

Serving provider

Input /1M

Output /1M

Endpoints

TencentCheapest

$0.435

$0.87

Standardtencent/deepseek-v4-pro

Component	Unit	Standard
Text cached in	/1M tok	$0.00363
Text cache write	/1M tok	$0.00
Text input	/1M tok	$0.435
Text output	/1M tok	$0.87

VercelVercel reseller quote

$0.66

$1.98

OpenRouter

$0.955

$1.91

Wandb

$1.15

$2.55

Fireworks AI

$1.20

DeepInfra

$1.30

$2.60

DeepSeek

$1.32

$3.96

Novita

$1.60

$3.20

Dashscope is dearest at $2.40 / $4.80

Price history

listed seller prices, input and output per 1M tokens

Input Output

Input down 75% since first tracked

Cost calculator

estimate your monthly spend on this model

Input tokens / call

Output tokens / call

Calls / month

Serving provider

$157

estimated / month

Model IDs

copy the exact identifier for your platform

tencent/deepseek-v4-prodeepseek/deepseek-v4-prodeepseek/deepseek-v4-prowandb/deepseek-ai/DeepSeek-V4-Profireworks_ai/accounts/fireworks/models/deepseek-v4-prodeepinfra/deepseek-ai/DeepSeek-V4-Prodeepseek-v4-pronovita/deepseek/deepseek-v4-proaihubmix/deepseek-v4-proazure_ai/deepseek-v4-protogether_ai/deepseek-ai/DeepSeek-V4-Probaseten/deepseek-ai/DeepSeek-V4-Pronebius/deepseek-ai/DeepSeek-V4-Prodashscope/deepseek-v4-proqwen_ai_platform/deepseek-v4-proqwencloud/deepseek-v4-pro

Frequently asked questions

DeepSeek V4 Pro 0423 pricing, context and availability

Is DeepSeek V4 Pro 0423 still available?

Yes. DeepSeek V4 Pro 0423 is scheduled for retirement on 2028-02-20.

How much does DeepSeek V4 Pro 0423 cost?

DeepSeek V4 Pro 0423 costs $0.435 per 1M input tokens from Tencent, and $0.87 per 1M output tokens from Tencent. Across 16 serving providers, input prices range from $0.435 to $2.40 per 1M tokens.