Skip to content
Leaders Board.
Models/DeepSeekOpen weights

DeepSeek V4.1 Flash

#33 of all models across the language-model boards.
#33
Overall rank
3.3
Podium score
0
Board wins
Sep 9, 2026
Released
·Reasoning·Coding2Speed6Price2Usage

Reasoning

Hard questions with checkable answers: science, math, puzzles, expert exams.

6
LiveBench
81.1% · of 59
as “deepseek-v4.1-flash-max”
18
AA Intelligence
39.5 · of 100
as “DeepSeek V4.1 Flash (Reasoning, Max Effort)”
28
Epoch ECI
155.0 · of 100
as “DeepSeek V4.1 Flash”

Coding

Fixing real repositories, building web apps, editing code.

15
WebDev Arena
1,620 · of 100
as “deepseek-v4.1-flash-max”

Speed

Output tokens per second on the vendor API.

2
AA Speed
232 tok/s · of 24
as “DeepSeek V4.1 Flash (Reasoning, Max Effort)”

Price

Blended API price per million tokens, cheapest first.

6
AA Price
$0.53 · of 27
as “DeepSeek V4.1 Flash (Reasoning, Max Effort)”

Usage

What developers actually send tokens to.

2
OpenRouter
18.4T tok · of 19
as “deepseek/deepseek-v4.1-flash-20260910”

More from DeepSeek

DeepSeek V4 Pro 0813DeepSeek V4 Flash VisionDeepSeek-V3.2DeepSeekDeepSeek V3DeepSeek-R1 (May 2025)DeepSeek V3.1 TerminusDeepSeek V3.1

Scores as published by each board on the capture date. Model names and logos belong to their owners; logos via logo.dev.

Weekly: the AI leaderboards with a new #1, Saturday mornings.