impulse

DeepSeek V4 Pro

Open weights

DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window.

API id deepseek-v4-proReleased 24 Apr 2026 (Apr 2026)
Input / 1M
$1.32
at DeepSeek
Output / 1M
$3.96
at DeepSeek
Context window
1M
tokens
Max output
393K
tokens per reply
Hosts
24
37 listings
Released
Apr 2026
24 Apr 2026

Pricing

List prices at DeepSeek, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$1.32$3.96
Cached input
97% off input
$0.044
Cache write, 5 minutes$0.00

What it costs a month

At DeepSeek list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$578
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$470
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$750
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$4,118
Use your own numbers

Where to run it

24 hosts serve DeepSeek V4 Pro. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
DeepSeekLab API
deepseek-v4-pro
$1.32 / $3.96
deepseek-v4-pro
$0.528 / $1.58
−60% vs headline
Fireworks AIInference
deepseek-v4-pro
$1.20 / $1.20
−39% vs headline
ai&Cloud
deepseek-v4-pro
$1.00 / $2.50
−31% vs headline
SailCloud
0813
DeepSeek-V4-Pro-0813
$0.92 / $2.77
−30% vs headline
W&B InferenceInference
DeepSeek-V4-Pro
$1.15 / $2.55
−24% vs headline
GMI CloudInference
0813
DeepSeek-V4-Pro-0813
$1.06 / $3.17
−20% vs headline
DeepInfraInference
DeepSeek-V4-Pro
$1.30 / $2.60
−18% vs headline
0813
databricks-deepseek-v4-pro-0813
$1.32 / $3.96
deepseek-v4-pro
$1.32 / $3.96
Together AIInference
0813
DeepSeek-V4-Pro-0813
$1.32 / $3.96
PerplexityGateway
0813
deepseek-v4-pro-0813
$1.32 / $3.96
Novita AIInference
deepseek-v4-pro
$1.60 / $3.20
+1% vs headline
deepseek-v4-pro
$1.65 / $3.30
+4% vs headline
AIHubMixGateway
deepseek-v4-pro
$1.69 / $3.38
+7% vs headline
deepseek-v4-pro
$1.74 / $3.48
+10% vs headline
deepseek-v4-pro
$1.74 / $3.48
+10% vs headline
Retires 20 Feb 2028
deepseek-v4-pro
$1.74 / $3.48
+10% vs headline
BasetenInference
DeepSeek-V4-Pro
$1.74 / $3.48
+10% vs headline
DeepSeek-V4-Pro
$1.75 / $3.50
+10% vs headline
DeepSeek-V4-Pro
$1.72 / $5.15
+30% vs headline
deepseek-v4-pro
$2.40 / $4.80
+52% vs headline
deepseek-v4-pro
$2.40 / $4.80
+52% vs headline
Qwen CloudLab API
deepseek-v4-pro
$2.40 / $4.80
+52% vs headline
Show 13 more listings: regions, snapshots and variants
DeepSeek
Alias
deepseek-v4-pro
$1.32 / $3.96
Fireworks AI
0813
deepseek-v4-pro-0813
$1.32 / $3.96
0813
deepseek-v4-pro-0813
$1.32 / $3.96
W&B Inference
0813
DeepSeek-V4-Pro-0813
$1.31 / $3.96
DeepInfra
0813
DeepSeek-V4-Pro-0813
$1.30 / $2.60
−18% vs headline
Novita AI
0813
deepseek-v4-pro-0813
$1.32 / $3.96
EmpirioLabs AI
0813
deepseek-v4-pro-0813
$1.32 / $3.96
StreamLake
0813
DeepSeek-V4-Pro-0813
$1.32 / $3.96
Tencent TokenHub
0813
deepseek-v4-pro-0813
$1.32 / $3.96
Baseten
0813
DeepSeek-V4-Pro-0813
$1.32 / $3.96
Nebius AI Studio
0813
DeepSeek-V4-Pro-0813
$1.32 / $3.96
Inference APIs
Alias
deepseek-v4-pro
$1.72 / $5.15
+30% vs headline
0813
DeepSeek-V4-Pro-0813
$1.72 / $5.15
+30% vs headline

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
Text
Produces
Text
Features
Assistant prefillParallel tool callsPrompt caching
Open weights
License: DeepSeek License (MIT code)
Runs in about 3837.2 GB of GPU memory
Runtimes: SGLang, TGI, vLLM

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

  • CLF AI Gateway
    deepseek-v4-pro
    $0.528 / $1.58
    60% less
  • Fireworks AI
    deepseek-v4-pro
    $1.20 / $1.20
    39% less
  • ai&
    deepseek-v4-pro
    $1.00 / $2.50
    31% less
  • Sail
    DeepSeek-V4-Pro-0813
    $0.92 / $2.77
    30% less
  • W&B Inference
    DeepSeek-V4-Pro
    $1.15 / $2.55
    24% less

Smaller DeepSeek models

No cheaper DeepSeek model keeps its key capabilities.

Other frontier labs

  • GPT-5.4 Mini
    OpenAI · on OpenAI
    $0.75 / $4.50
    Has every key capability
    15% less
  • Gemini 3.8 Flash
    Google · on Google AI Studio
    $0.75 / $3.75
    Has every key capability
    24% less
  • Grok Build 0.1
    xAI · on xAI
    $1.00 / $2.00
    Has every key capability
    37% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    90% less

Open-weight picks

  • Inkling
    Independent · on DeepInfra
    $0.95 / $4.05
    Has every key capability
    13% less
  • GLM 5.3
    Zhipu AI · on CLF AI Gateway
    $0.56 / $1.76
    Has every key capability
    57% less
  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    65% less
  • DeepSeek Flash
    DeepSeek · on DeepSeek
    $0.30 / $1.20
    Has every key capability
    73% less

Call it

Using DeepSeek's API.

Python
from litellm import completion

response = completion(
    model="deepseek/deepseek-v4-pro",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does DeepSeek V4 Pro cost?

On DeepSeek, DeepSeek V4 Pro costs $1.32 per million input tokens and $3.96 per million output tokens. Cached input costs $0.044 per million tokens. The cheapest host is CLF AI Gateway, at $0.528 per million input tokens and $1.58 per million output tokens.

Which APIs offer DeepSeek V4 Pro?

DeepSeek V4 Pro runs on 24 APIs: DeepSeek, CLF AI Gateway, Fireworks AI, ai&, Sail, W&B Inference, GMI Cloud, DeepInfra, Databricks and 15 more.

What is DeepSeek V4 Pro's context window?

1M tokens, with up to 393K tokens in one reply.

Is DeepSeek V4 Pro being retired?

Azure AI Foundry retires it on 20 Feb 2028. 23 hosts keep serving it.

What are cheaper alternatives to DeepSeek V4 Pro?

Models that cost less and keep most of what it can do: GPT-5.4 Mini on OpenAI (15% less), Gemini 3.8 Flash on Google AI Studio (24% less) and Inkling on DeepInfra (13% less). DeepSeek V4 Pro itself costs 60% less on CLF AI Gateway. Savings are on a mix of 3 input tokens to 1 output token.

When was DeepSeek V4 Pro released?

DeepSeek released it on 24 Apr 2026.

What is DeepSeek V4 Pro's API model id?

Use deepseek-v4-pro with DeepSeek. Other hosts use their own ids: DeepSeek-V4-Pro-0813 on Sail, DeepSeek-V4-Pro on W&B Inference, DeepSeek-V4-Pro-0813 on GMI Cloud, DeepSeek-V4-Pro on DeepInfra.

Sources

Raw catalog entry for deepseek-v4-pro on DeepSeek
{
  "cache_creation_input_token_cost": 0,
  "cache_read_input_token_cost": 4.4e-8,
  "input_cost_per_token": 0.00000132,
  "input_cost_per_token_cache_hit": 4.4e-8,
  "litellm_provider": "deepseek",
  "max_input_tokens": 1000000,
  "max_output_tokens": 393216,
  "mode": "chat",
  "off_peak_pricing": {
    "cache_read_input_token_cost": 2.2e-8,
    "input_cost_per_token": 6.6e-7,
    "output_cost_per_token": 0.00000198,
    "windows": [
      {
        "hours_utc": [
          "00:00-01:00",
          "04:00-06:00",
          "10:00-00:00"
        ],
        "weekdays": [
          1,
          2,
          3,
          4,
          5
        ]
      },
      {
        "hours_utc": "00:00-00:00",
        "weekdays": [
          6,
          7
        ]
      }
    ]
  },
  "output_cost_per_token": 0.00000396,
  "source": "https://api-docs.deepseek.com/quick_start/pricing",
  "supported_endpoints": [
    "/v1/chat/completions"
  ],
  "supports_assistant_prefill": true,
  "supports_function_calling": true,
  "supports_native_streaming": true,
  "supports_parallel_function_calling": true,
  "supports_prompt_caching": true,
  "supports_reasoning": true,
  "supports_response_schema": true,
  "supports_system_messages": true,
  "supports_tool_choice": true,
  "supports_vision": false
}