ModelPriceWatch.com
Last scan 2026-08-22 Models tracked 220 Providers 31 Cheapest paid Granite 4.0 H Micro $0.017/Mtok in Every price links to its source

DeepSeek V4 Pro

by DeepInfra

Current mid tier mid tier

Today's price · per 1M tokens

Input

$1.30

per 1M tokens

Output

$2.60

per 1M tokens

Blended

$1.63

blended $/1M — 3:1 weighted input:output

Cached input

$0.100

8% of input — prompt caching

Source: official DeepInfra pricing · read Aug 14, 2026 MODELPRICEWATCH.COM · 2026-08-22

Price receipt

We read DeepInfra's own pricing page on Aug 17, 2026 and found DeepSeek-V4-Pro listed with a price on the same row — that read is where the number above comes from. We keep the page text we read; its content hash is 7f4b817fe1.

DeepSeek V4 Pro is available on 4 hosts. See the full cheapest-first comparison on the consolidated DeepSeek V4 Pro page →

Overview

DeepSeek V4 Pro served on DeepInfra at $1.30/$2.60 per 1M tokens, with cached input at $0.10 and a 1M-token context. Undercuts the other tracked third-party hosts of this model (Fireworks and Together, both $1.74/$3.48) on both input and output.

Capabilities

struck through = not supported
Input 1/5
Text ✓ Image Audio Video PDF
Output 1/5
Text ✓ Image Audio Video Embedding
Features 3/9
Prompt caching ✓ Reasoning ✓ Coding Fast inference Long context ✓ Open weights Multimodal Web search Realtime

Benchmark performance

accuracy % · higher is better
Avg benchmark score73.5%
Perf per $/Mtok45.2
GPQA Diamond
90.1%
SWE-Bench Verified vendor
80.6%
LiveCodeBench
93.5%
MCP Atlas
73.6%
BrowseComp
83.4%
Humanity's Last Exam
48.2%
ARC-AGI 2
30%
AIME 2025
75%
MMMLU
82%
BFCL
72%
HumanEval
80.6%
MATH 500
80%
Independent composite scores
  • AA Intelligence Index: 53.2 (Artificial Analysis composite across reasoning, knowledge and coding evals)
  • AA Agentic Index: 49.6 (Tool use, planning, autonomy)
  • AA-Omniscience Index: 0.8 (−100–100)
  • GDPval-AA v2: 1590.4 ELO (Real-world work tasks, human baseline = 1000)
How it stacks up
  • Ranks #30 of 69 benchmarked models by average score
  • Ranks #37 of 161 comparably-measured models by percentile score, across 15 independent measurements
  • Ranks #61 of 161 comparably-tested models by normalized performance per dollar
  • Strongest at LiveCodeBench — 93.5%, #4 of 25

174.9 tokens/sec output

Source: Vellum LLM Leaderboard · updated Aug 14, 2026 · See full rankings →

Specifications

Provider
DeepInfra
Context window
1M tokens
Modality
text
Parameters
Proprietary
Open source
No — proprietary
Released
Apr 24, 2026
Status
Current
Last updated
Aug 14, 2026
Tags
reasoningcachingmid-tier

Availability verified: Aug 17, 2026 — listed on DeepInfra's own page