DeepInfra raises $107M Series B to scale the inference cloud — read the announcement

Best DeepSeek V4 Pro API Providers (2026): Price & Speed Compared

Published on 2026.05.25 by DeepInfra
Best DeepSeek V4 Pro API Providers (2026): Price & Speed Compared

Updated November 2026 for the DeepSeek-V4-Pro-0813 release.

DeepSeek V4 Pro reached general availability on August 13, 2026 as DeepSeek-V4-Pro-0813. It keeps the preview’s architecture: a Mixture-of-Experts model with 1.6T total and 49B active parameters, a 1M-token context window and up to 384K output tokens. On top of that it adds a built-in DSpark speculative decoding module and a three-level reasoning_effort control (low, high, max). The 0813 model card shows big vendor-reported gains on agentic and coding tasks. DeepSWE, for example, rose from 12.8 to 62.7. Because the weights are MIT-licensed, a lot of providers serve it, and they differ on price, speed and how long you wait for the first useful token.

This guide compares the providers that serve the 0813 build, based on pricing pulled on 2026-11-16. Speed, latency, blended price and error rate come from Artificial Analysis, which measures the model in its Max reasoning mode. Routed prices come from OpenRouter, and rates and features come from each provider’s own model or pricing page. V4 Pro reasons before it answers, so this guide tracks two latency figures. Time to first token (TTFT) measures when the stream starts. Time to first answer token measures when the model has finished thinking and starts writing the answer. For this model, every provider has a TTFT under 3 seconds, but the first answer token arrives anywhere from 32 to 74 seconds after the request, depending on the provider. The second figure is the one your users feel.

Summary of the Best DeepSeek V4 Pro API Providers

ProviderBest For
DeepInfraLowest price overall: lowest output price on Artificial Analysis, a Flex tier at $2.08 output, 0% measured error rate
DeepSeek (official API)First-party access, off-peak discounts and the cheapest cache reads
DigitalOceanFastest output and lowest latency to the first answer token
BasetenSecond-fastest output speed on a per-token Model API
Nebius Token FactoryOutput speed on par with DeepSeek’s API, with dedicated endpoints
Novita AI0% measured error rate with a low cost per agentic task
SiliconFlowFast TTFT, plus on-demand and reserved GPU options
OpenRouterOne API key, provider fallback and access to 20 providers

Measured price and performance (Max reasoning, pulled 2026-11-16)

Prices are per 1M tokens. AA’s blended price weights cache hits, input and output 7:2:1. Speed, latency and error rate are Artificial Analysis headline values.

ProviderInputOutputCache hitBlended (7:2:1)Output speedTTFTFirst answer tokenError rate
DeepInfra$1.30$2.60$0.10$0.5965.1 t/s2.17 s67.09 s0.0%
DeepSeek (peak)$1.32$3.96$0.044$0.6993.5 t/s2.57 s52.73 s0.8%
DigitalOcean$1.32$3.96$0.044$0.69140.0 t/s1.72 s31.67 s3.1%
Baseten$1.32$3.96$0.132$0.75114.2 t/s2.84 s43.87 s0.8%
Nebius$1.32$3.96$0.04$0.6993.9 t/s2.92 s45.40 s0.8%
Novita$1.32$3.96$0.044$0.6977.9 t/s2.49 s59.31 s0.0%
SiliconFlow$1.32$3.96$0.044$0.6941.4 t/s1.72 s73.93 s0.9%

Pricing as of 2026-11-16. DeepSeek’s first-party prices are halved off-peak (see below).

Detailed Provider Reviews

1. DeepInfra

DeepInfra is the best pick when cost per token is the deciding factor. Its output price is the lowest that Artificial Analysis measures for V4 Pro 0813, and its Flex tier goes lower still for workloads that can tolerate queueing.

  • Model ID: deepseek-ai/DeepSeek-V4-Pro-0813, served in fp8
  • Standard rates: $1.30 input / $2.60 output / $0.10 cached input per 1M tokens (no promo; this is list price)
  • Flex tier: $1.04 / $2.08 / $0.08. Opt in with "service_tier": "flex". A request waits up to 10 minutes for capacity, then returns HTTP 429, and a rejected Flex request is not charged (how Flex works)
  • Priority tier: $1.95 / $3.90 / $0.15 for latency-sensitive traffic ("service_tier": "priority")
  • Blended price (AA, 7:2:1): $0.59 at the standard rate, below every other provider in this guide ($0.69–$0.75). With the same weighting, Flex works out to about $0.47
  • Measured performance (AA, Max): 65.1 t/s output, 2.17 s TTFT, 67.09 s to first answer token, 0.0% error rate, 98.3% cache hit rate
  • Context and output: 1M-token context; up to 384K output tokens (393,216). If max_tokens is unset, the default is the lesser of the remaining context and 65,536
  • Features: OpenAI-compatible API, tool/function calling, JSON mode, structured output, reasoning_effort from none (thinking off) to max, prompt caching with prompt_cache_key, and up to 4 fallback models per request
  • Platform: usage-based pricing with no minimums, zero data retention, SOC 2 and ISO 27001, private deployments
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["DEEPINFRA_API_KEY"],
    base_url="https://api.deepinfra.com/v1/openai",
)

resp = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V4-Pro-0813",
    messages=[{"role": "user", "content": "Refactor this function to remove the N+1 query."}],
    extra_body={"reasoning_effort": "high", "service_tier": "flex"},
)
print(resp.choices[0].message.content)
copy

DeepInfra fits teams running V4 Pro at volume: agent backends, batch coding jobs, evaluation runs, and long-context document work where output tokens and reasoning tokens (billed as output) drive the bill. Flex is useful for overnight and asynchronous pipelines, and Priority covers the requests that can’t wait. DeepInfra does not have the cheapest cache reads. Several providers charge $0.044 or less, so prompts that are almost entirely cache hits can cost less elsewhere. And if response speed is your product, test DigitalOcean. The original deepseek-ai/DeepSeek-V4-Pro ID is still available at the same price, but new integrations should use the 0813 ID. Full cost scenarios are in the DeepSeek V4 Pro pricing guide.

2. DeepSeek (official API)

The DeepSeek API is the best pick when you want first-party access, can schedule work into off-peak hours, or run cache-heavy prompts.

  • Model name: deepseek-v4-pro, which now serves DeepSeek-V4-Pro-0813
  • Peak rates: $1.32 input (cache miss) / $3.96 output / $0.044 cache hit
  • Off-peak rates (50% of peak): $0.66 / $1.98 / $0.022. Peak hours are 01:00–04:00 and 06:00–10:00 UTC, Monday to Friday, excluding Chinese public holidays. All other hours, including weekends, are off-peak
  • Measured performance (AA, Max): 93.5 t/s, 2.57 s TTFT, 52.73 s to first answer token, 0.8% error rate, and the lowest cost per Intelligence Index task of any provider ($1.77)
  • Context and output: 1M context, 384K maximum output
  • Features: JSON output, tool calls, OpenAI Responses API and Anthropic API formats, chat prefix completion (beta), FIM completion in non-thinking mode. No vision on V4 Pro
  • Limits: concurrency limit of 500 for deepseek-v4-pro

The first-party API has the lowest cache-hit price in this guide, and off-peak its output rate ($1.98) is below DeepInfra’s standard $2.60. During peak hours it costs more than DeepInfra on output ($3.96 vs $2.60). It’s a strong fit for teams that can batch work into off-peak windows or that want features only DeepSeek offers, such as the Responses API. The trade-offs are time-of-day pricing that makes costs harder to forecast and a 500-request concurrency cap.

3. DigitalOcean

DigitalOcean Serverless Inference is the best pick when speed and latency are the product. It leads Artificial Analysis on output speed and on time to first answer token for V4 Pro 0813.

  • Rates: $1.32 input / $3.96 output / $0.044 cache read per 1M tokens
  • Blended price (AA): $0.69
  • Measured performance (AA, Max): 140.0 t/s output (fastest), 1.72 s TTFT (tied lowest), 31.67 s to first answer token (lowest), 11.8 min per Index task (fastest), 3.1% error rate
  • Access: Serverless Inference and Inference Router, prepaid billing; dedicated GPU inference is billed per GPU-hour

DigitalOcean is the provider to test for chat UIs, interactive coding assistants, and any agent loop where a user is waiting for the answer. Its first answer token arrives up to 42 seconds sooner than the slowest provider in this guide. It costs more per output token than DeepInfra ($3.96 vs $2.60), and its 3.1% error rate is the highest Artificial Analysis measured for this model, so build in retries.

4. Baseten

Baseten’s Model API is the best pick when you want high output speed on per-token pricing, with the option to move to dedicated deployments later.

  • Rates: $1.32 input (cache miss) / $3.96 output / $0.132 cache hit
  • Blended price (AA): $0.75, the highest in this guide because of its cache rate
  • Measured performance (AA, Max): 114.2 t/s (second fastest), 2.84 s TTFT, 43.87 s to first answer token (second lowest), 0.8% error rate
  • Context: 1M
  • Features: tool use, JSON mode, structured outputs, reasoning; the same library offers V4 Flash 0731 and V4.1 Flash for sub-agents

Baseten suits teams building coding agents who want fast generation and a single vendor for a Pro main agent plus Flash sub-agents. Its cache-hit price is 3× DeepSeek’s and above DeepInfra’s $0.10, so long cached contexts cost more here.

5. Nebius Token Factory

Nebius Token Factory is the best pick when you want output speed on par with DeepSeek’s own API, with dedicated endpoints as an option.

  • Rates (AA): $1.32 input / $3.96 output / $0.04 cache hit
  • Blended price (AA): $0.69
  • Measured performance (AA, Max): 93.9 t/s, 2.92 s TTFT, 45.40 s to first answer token, 0.8% error rate
  • Model lifecycle: Nebius has deprecated the serverless deepseek-ai/DeepSeek-V4-Pro ID and recommends deepseek-ai/DeepSeek-V4-Pro-0813 as its replacement. Requests aren’t rerouted automatically, and dedicated endpoints are unaffected

Nebius’s speed and first-answer latency are close to DeepSeek’s own API. The catch is cost on agentic workloads. Artificial Analysis puts its cost per Index task at $13.43 with a 54.9% cache hit rate, far above every other provider here. If your workload depends on prompt caching, test your cache hit rate before committing.

6. Novita AI

Novita AI is the best pick when reliability and cost per agentic task matter more than raw speed.

  • Model ID: deepseek/deepseek-v4-pro-0813, serverless, OpenAI-compatible
  • Rates (AA): $1.32 input / $3.96 output / $0.044 cache hit. Novita’s fp8 route on OpenRouter lists $0.99 / $2.97 / $0.033
  • Blended price (AA): $0.69
  • Measured performance (AA, Max): 77.9 t/s, 2.49 s TTFT, 59.31 s to first answer token, 0.0% error rate, $1.86 per Index task (second lowest)

Novita and DeepInfra are the two providers with a 0% error rate in Artificial Analysis’s runs. Novita’s cheap cache reads give it a low cost per agentic task. Its output price is higher than DeepInfra’s at either rate ($2.97 via OpenRouter, $3.96 list vs $2.60).

7. SiliconFlow

SiliconFlow is the best pick when you want a fast stream start and the option to move from serverless to dedicated GPUs.

  • Model ID: deepseek-ai/DeepSeek-V4-Pro-0813
  • Rates (AA): $1.32 input / $3.96 output / $0.044 cache hit
  • Blended price (AA): $0.69
  • Measured performance (AA, Max): 41.4 t/s, 1.72 s TTFT (tied lowest), 73.93 s to first answer token, 0.9% error rate
  • Deployment: serverless, on-demand dedicated deployment, and reserved GPUs; JSON mode and structured outputs; no image input

SiliconFlow’s low TTFT shows how misleading that metric is for a reasoning model. The stream starts as fast as DigitalOcean’s, but the answer arrives last of the measured providers. It makes sense if you are moving toward reserved capacity, or already use SiliconFlow for other DeepSeek models. It is not the pick for interactive use on the serverless tier.

8. OpenRouter

OpenRouter is the best pick when you want one API key across providers, with automatic fallback when an endpoint degrades.

  • Coverage: 26 endpoints from 20 providers on the 0813 slug, including DeepInfra’s standard ($1.30 / $2.60), Flex ($1.04 / $2.08) and Priority ($1.95 / $3.90) endpoints
  • Price range: input from $0.10 to $1.98 and output from $1.742 to $8.00 per 1M tokens, depending on provider and quantization
  • Quantization varies: endpoints are listed as fp8, fp4, mxfp4 or unknown, so pin a provider if output consistency matters

OpenRouter is a good fit for prototyping and for production apps that need uptime insurance across several backends. It lists a few routes with lower output prices than DeepInfra’s standard rate, such as Baidu at $1.976, plus DeepSeek’s off-peak price. Some of those routes show degraded status or an unstated quantization, and routing means you don’t always control which backend serves a given request. For the DeepInfra rates without a router in between, call DeepInfra directly.

Using V4 Flash or V4.1 Flash instead

This post originally covered both V4 variants. DeepSeek’s smaller models have moved on since May: the original V4 Flash was followed by a GA build (V4 Flash 0731) and then by V4.1 Flash. On Artificial Analysis’s Intelligence Index, V4.1 Flash 1017 scores 33, against 23 for V4 Pro 0813 (Max). On DeepSeek’s own API, the legacy deepseek-v4-flash name is now served by V4.1 Flash 1017. If your workload doesn’t need V4 Pro specifically, a Flash model is much cheaper. Current DeepInfra prices per 1M tokens:

Model on DeepInfraInputOutputCached input
DeepSeek-V4.1-Flash-1017 (51% off list)$0.098$0.294$0.0029
DeepSeek-V4.1-Flash (30% off list)$0.14$0.42$0.0042
DeepSeek-V4-Flash-0731$0.06$0.18$0.015
DeepSeek-V4-Flash$0.09$0.18$0.018

Both V4.1 Flash models list at $0.20 / $0.60 / $0.006 before the promotion. All four support Flex and Priority and have a 1M-token context.

Final Thoughts

The right DeepSeek V4 Pro provider depends on which constraint binds first:

  • Lowest price per token: DeepInfra. It has the lowest output price on Artificial Analysis ($2.60), a $0.59 blended price, Flex at $1.04 / $2.08 for async work, and a 0% measured error rate.
  • Fastest answers for users: DigitalOcean (140.0 t/s, 31.67 s to first answer token), then Baseten (114.2 t/s, 43.87 s).
  • First-party access and off-peak batch work: DeepSeek’s API ($0.66 / $1.98 off-peak) for the Responses API and cheap cache reads, provided you can live with time-of-day pricing.
  • Cache-dominated agent loops: compare cost per task, not list price. DeepSeek ($1.77) and Novita ($1.86) lead Artificial Analysis on that measure.
  • Dedicated or reserved capacity: SiliconFlow, Nebius or Baseten.
  • One key and fallback across providers: OpenRouter.

The Best Overall Solution

For most production workloads, DeepInfra is the best overall choice for DeepSeek V4 Pro 0813. It has the lowest output price that Artificial Analysis measures, the cheapest Flex tier for anything that can wait, and Priority for requests that can’t. It also gives you an OpenAI-compatible API with the full 1M context and 384K output, and it posted zero errors in Artificial Analysis’s runs. Output tokens, including reasoning tokens, usually make up most of a V4 Pro bill, so that output price matters more than any other rate on the page.

Start with the model page or go straight to the API reference. For lighter workloads, DeepSeek V4.1 Flash costs a fraction as much. The DeepSeek V4 Pro pricing guide covers workload cost scenarios, and the API benchmarks guide has the full speed and latency comparison. Early-stage startups can apply for free credits through DeepStart.

Related articles
Qwen3.5 9B API Benchmarks: Latency, Throughput & CostQwen3.5 9B API Benchmarks: Latency, Throughput & Cost<p>About Qwen3.5 9B Qwen3.5 9B is the flagship of Alibaba&#8217;s Qwen3.5 Small Model Series, released on March 2, 2026. It is a dense multimodal model combining Gated Delta Networks (a form of linear attention) with a sparse Mixture-of-Experts system, enabling higher throughput and lower latency during inference compared to traditional dense architectures. The architecture utilizes [&hellip;]</p>
Hosted Agents: your own always-on AI agent, from $13/monthHosted Agents: your own always-on AI agent, from $13/monthOne click gives you a dedicated, isolated AI agent, pre-wired to fast inference and ready to work the moment it boots. No VMs, no SSH hardening, no patching. From $13/month, and idle is free.
Kimi K3.1 Pro Pricing: API Costs, Providers & Flex (2026)Kimi K3.1 Pro Pricing: API Costs, Providers & Flex (2026)<p>Kimi K3.1 Pro Executive Summary Pricing as of 2026-11-17. DeepInfra&#8217;s Standard rate is promotional (5% off a $2.00 / $10.00 list price, no published end date). Budget against list if you&#8217;re planning past this quarter. Best For Provider Recommendation Why Lowest price with a 0% error rate DeepInfra Standard $1.90 input / $9.50 output / [&hellip;]</p>