Skip to main content

DeepSeek V4 Flash vs DeepSeek V4 Pro

SpecDeepSeek V4 FlashDeepSeek V4 Pro
Respan Index45.4 (#54)49.2 (#44)
Blended price per 1M$0.113$0.585
Context window1M1M
Input / output per 1M$0.09 / $0.18$0.28 / $1.5
Max output-384K
Inputtexttext
Released2026-07-312026-08-13

Respan Index by category

Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.

CategoryDeepSeek V4 FlashDeepSeek V4 Pro
Coding29.7 DeepSeek V4 Flash29.8 DeepSeek V4 Pro
Agents58.0 DeepSeek V4 Flash59.7 DeepSeek V4 Pro
Reasoning46.1 DeepSeek V4 Flash52.8 DeepSeek V4 Pro
Knowledge & language35.5 DeepSeek V4 Flash52.6 DeepSeek V4 Pro
Human preference55.9 DeepSeek V4 Flash56.0 DeepSeek V4 Pro

How the Respan Index works

Benchmarks both report

DeepSeek V4 Pro leads on 17 of the 22 benchmarks both report. Higher is better. Each score links to where it was published.

BenchmarkDeepSeek V4 FlashDeepSeek V4 Pro
LMArena Elo14361464
LMArena WebDev (LMArena)15821581
Chess Puzzles (Epoch AI)32%46%
AIME 202695.8%96.7%
DeepSWE v1.1 (Datacurve)53.3%62.8%
FrontierMath Tier 4 v2 (Epoch AI)24.4%26.8%
FrontierMath Tiers 1-3 v2 (Epoch AI)57.5%64.6%
GPQA Diamond91%91.7%
LiveBench Agentic Coding (LiveBench)46.8%54.9%
LiveBench Coding (LiveBench)75%77.2%
LiveBench Data Analysis (LiveBench)79.3%79.2%
LiveBench Instruction Following (LiveBench)65.5%67.7%
LiveBench Language (LiveBench)79.2%82.1%
LiveBench Mathematics (LiveBench)86.8%95.1%
LiveBench Reasoning (LiveBench)86.6%85.8%
Mystery Game Puzzles (Epoch AI)34%43%
OTIS Mock AIME 2024-2025 (Epoch AI)94.4%98.6%
SimpleQA Verified33.6%52.9%
Terminal-Bench 4.0 (Vals AI)18.7%14.1%
WeirdML (Håvard Tveit Ihle)63%66.2%
SimpleBench (SimpleBench)61.1%50.9%
Toolathlon-Verified (HKUST)70.7%74.4%

FAQ

Is DeepSeek V4 Flash better than DeepSeek V4 Pro?

DeepSeek V4 Pro scores 49.2 on the Respan Index (#44), ahead of DeepSeek V4 Flash at 45.4 (#54). DeepSeek V4 Pro leads on 17 of the 22 benchmarks both report.

Which is cheaper, DeepSeek V4 Flash or DeepSeek V4 Pro?

DeepSeek V4 Flash lists at $0.09 input and $0.18 output per 1M tokens; DeepSeek V4 Pro at $0.28 and $1.5. DeepSeek V4 Flash is 5.2x cheaper per token at list price.

Which has the larger context window?

Both accept 1M tokens of context.

Compare other models

vs

Built for AI agents.
Break less.
Ship more.