AI BENCHY
Advertise here
#140

DeepSeek V4 Flash

DeepSeek Release: 2026-04-24 Tested on: 2026-06-04 13:02 deepseek/deepseek-v4-flash::none
(high) (none)

Summary

DeepSeek V4 Flash scores 5.0 on AI BENCHY and ranks #140. It has 10.0 reliability, a 30.2% pass rate, $0.008 total cost, and 26.72s average response time.

What makes DeepSeek V4 Flash unique: It stands out most in Data parsing and extraction, where it ranks #1, while Anti-AI Tricks is its weakest area at #12.

Score

5.0

Consistency

8.9

Reliability

10.0

Total Cost (Current Price)

$0.008 ↓ -25.1%

Tested at: $0.010

Total Output Tokens

13,708

Total Input Tokens

48,401

Input Price

$0.099 / 1M

Output Price

$0.197 / 1M

Tests Correct

Wrong Tests: 16

Attempt pass rate: 30.2%

Flaky tests

3

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

26.72s

Response Time (max): 111.96s

Response Time (total): 561.20s

Wrong answer: 11 Extra formatting: 3 Did not follow instructions: 1 Invalid tool call: 1

Generation showcase

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#140 DeepSeek V4 Flash

none
Cost
$0.004
Time
157.6s
Tokens
11,297 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-06-04 14:22 Re-test 5.0 10.0 $0.008 ↓ Compare
2026-06-04 13:02 New test added 5.0 10.0 $0.008 ↓ Current run
2026-05-22 00:37 Suite changed 5.1 10.0 $0.010 Compare
2026-04-24 09:19 Initial run 5.3 N/A $0.008 Compare

This run used a different benchmark suite. Keep suite changes in mind when reading historical movement.

Price History

Historical pricing data for this model from OpenRouter.

Date Input Price Output Price
2026-06-04 13:03 $0.099 / 1M $0.197 / 1M

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

DeepSeek V4 FlashnonevsGLM 5.1noneDeepSeek V4 FlashnonevsQwen3.5 Plus 2026-04-20noneDeepSeek V4 FlashnonevsQwen3.5-27BnoneDeepSeek V4 FlashnonevsSeed-2.0-LitenoneDeepSeek V4 FlashnonevsGemini 3 Flash PreviewmediumDeepSeek V4 FlashnonevsGemini 3.5 FlashhighDeepSeek V4 FlashnonevsNemotron 3 Ultra 550b A55bmediumFree AvailableDeepSeek V4 FlashnonevsGemini 3.5 Flashlow

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 3.0 10.0
Coding 4.2 7.4
Combined 4.5 2.1
Data parsing and extraction 10.0 10.0
Domain specific 5.3 10.0
General Intelligence 4.2 9.9
Instructions following 6.5 10.0
Puzzle Solving 3.1 7.3
Tool Calling 10.0 10.0
Trivia 3.0 10.0

Compared models