Question 1

What is DeepSeek V4 Pro?

Accepted Answer

DeepSeek V4 Pro is a model from Deepseek, released on Apr 24, 2026. 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. It is designed for advanced reasoning, coding, and long-horizon agent workflows, with strong performance across knowledge, math, and software engineering benchmarks. Built on the same architecture as DeepSeek V4 Flash, it introduces a hybrid attention system for efficient long-context processing. Reasoning efforts high and xhigh are supported; xhigh maps to max reasoning. It is well suited for complex workloads such as full-codebase analysis, multi-step automation, and large-scale information synthesis, where both capability and efficiency are critical.

Question 2

How much does DeepSeek V4 Pro cost?

Accepted Answer

On the Respan gateway, input starts at $0.21 per 1M tokens on Relace, and output starts at $0.42 per 1M tokens on OpenRouter, after gateway discounts. Averaged across 28 providers, it costs $1.24 per 1M input tokens and $3.33 per 1M output tokens.

Question 3

Which providers serve DeepSeek V4 Pro on Respan?

Accepted Answer

28 providers serve DeepSeek V4 Pro: Relace, OpenRouter, Wafer, Deepseek, Nextbit256, Novita, and 11 more. Perplexity, TogetherAI, and Tera require your own provider API key (BYOK).

Question 4

What is DeepSeek V4 Pro's context window?

Accepted Answer

DeepSeek V4 Pro accepts up to 1.0M tokens of context on its largest provider route.

Question 5

How do I use DeepSeek V4 Pro with the Respan gateway?

Accepted Answer

Point any OpenAI-compatible SDK at https://api.respan.ai/api/ with your Respan API key and set the model to "deepseek/deepseek-v4-pro". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

Question 6

How fast is DeepSeek V4 Pro?

Accepted Answer

Over the last 30 days of requests through the Respan gateway, the fastest provider was Deepseek at 71 tokens per second, and the lowest response latency was 0.3 seconds on DeepInfra. Speed varies with request size and provider.

Question 7

Where does DeepSeek V4 Pro rank on the Respan Index?

Accepted Answer

DeepSeek V4 Pro scores 46.9 out of 100 on the Respan Index v1.1, ranking #29 of 138 models. By category: Coding 29.4, Agents 59.6, Reasoning 44.8, Knowledge & language 52.6, and Human preference 54.7. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by llmmetric.com.

Question 8

How does DeepSeek V4 Pro score on benchmarks?

Accepted Answer

DeepSeek V4 Pro has 33 published benchmark scores, including Terminal-Bench 2.1 87.9%, Toolathlon-Verified 74.4%, and GPQA Diamond 91.7%. Each score links to where it was published; the data is collected by llmmetric.com.

Provider	Context	Input	Output
Relace`relace/deepseek-ai/DeepSeek-V4-Pro`	1.0M	$0.21/M	$4.20/M
OpenRouter`openrouter/deepseek/deepseek-v4-pro`	1.0M	$0.21/M	$0.42/M
Relace`relace/deepseek-ai/DeepSeek-V4-Pro-0813`	1.0M	$0.21/M	$4.20/M
Wafer`wafer/DeepSeek-V4-Pro`	1.0M	$0.39/M	$3.49/M
Deepseek`deepseek/deepseek-v4-pro`	1.0M	$0.43/M	$0.87/M
OpenRouter`openrouter/deepseek/deepseek-v4-pro-0813`	1.0M	$0.66/M	$1.98/M
Nextbit256`nextbit256/deepseek:v4-pro-0813`	1.0M	$1.06/M	$3.17/M
NovitaSave 10%`novita/deepseek/deepseek-v4-pro-0813`	1.0M	$1.19/M	$3.56/M
DeepInfra`deepinfra/deepseek-ai/DeepSeek-V4-Pro`	1.0M	$1.30/M	$2.60/M
DeepInfra`deepinfra/deepseek-ai/DeepSeek-V4-Pro-0813`	1.0M	$1.30/M	$2.60/M
Baseten`baseten/deepseek-ai/DeepSeek-V4-Pro-0813`	1.0M	$1.32/M	$3.96/M
EmpirioLabs`empiriolabs/deepseek-v4-pro-0813`	1M	$1.32/M	$3.96/M
Inference.net`inference_net/deepseek-v4-pro-0813`	1.0M	$1.32/M	$3.96/M
Nebius`nebius/deepseek-ai/DeepSeek-V4-Pro-0813`	979K	$1.32/M	$3.96/M
Parasail`parasail/deepseek-ai/DeepSeek-V4-Pro-0813`	1.0M	$1.32/M	$3.96/M
Parasail`parasail/parasail-deepseek-v4-pro-0813`	1.0M	$1.32/M	$3.96/M
PerplexityBYOK only`perplexity/deepseek-v4-pro-0813`	1.0M	$1.32/M	$3.96/M
TogetherAIBYOK only`together_ai/deepseek-ai/DeepSeek-V4-Pro-0813`	1.0M	$1.32/M	$3.96/M
NovitaSave 10%`novita/deepseek/deepseek-v4-pro`	1.0M	$1.44/M	$2.88/M
Featherless`featherless_ai/deepseek-ai/DeepSeek-V4-Pro`	262K	$1.60/M	$3.20/M
EmpirioLabs`empiriolabs/deepseek-v4-pro`	1M	$1.65/M	$3.30/M
Nextbit256`nextbit256/deepseek:v4-pro`	1.0M	$1.74/M	$3.48/M
Parasail`parasail/deepseek-ai/DeepSeek-V4-Pro`	1.0M	$1.74/M	$3.48/M
Parasail`parasail/parasail-deepseek-v4-pro`	1M	$1.74/M	$3.48/M
TeraBYOK only`tera/deepseek-ai/DeepSeek-V4-Pro`	1.0M	$1.74/M	$3.48/M
Nebius`nebius/deepseek-ai/DeepSeek-V4-Pro`	1.0M	$1.75/M	$3.50/M
IO.net`ionet/deepseek-ai/DeepSeek-V4-Pro`	1.0M	$1.75/M	$3.54/M
Inference.net`inference_net/deepseek-v4-pro`	1.0M	$2.10/M	$4.20/M

DeepSeek V4 Pro

Providers

Certified where it counts.

Pricing

Cost by volume

Throughput and latency

Output speed

Time to first token

Response latency

Benchmarks

Respan Indexv1.1

Coding8

Apps

DeepSeek V4 Pro FAQ

Works with your entire stack

Built for AI agents.
Break less.
Ship more.

Agents4

Reasoning7

Math7

Knowledge4

Arena3