Advertise here
#218

Hy4 flash

Tencent Release: 2026-11-17 Tested on: 2026-11-17 19:07 tencent/hy4-flash::medium
~190B total (~12B active)MoEClosedEstimated
(medium) (none)

Hy4 Flash beats Hy4 in our tests for about 1/24 of the cost →

Summary

Hy4 flash scores 7.3 on AI BENCHY and ranks #218. It has 10.0 reliability, a 72.5% pass rate, $0.098 total cost, and 49.04s average response time.

What makes Hy4 flash unique: It stands out most in Domain specific, where it ranks #2, while Puzzle Solving is its weakest area at #16. It uses unusually many reasoning tokens, which can help explain its slower or more expensive runs.

Model facts

Researched on 2026-11-17

Estimated
Parameters
~190B total (~12B active)
Architecture
MoE
Availability
Closed
License
-

Best estimate from public evidence; the vendor did not disclose every value. OpenRouter describes Hy4 flash as served with roughly a quarter of the flagship Hy4 parameter count. Estimates scale the Hy4 Preview backbone (770B total / 49B active MoE) by one quarter; Tencent has not published exact counts, weights or a license for this checkpoint. Fallback date records this onboarding request date.

Score

7.3

Consistency

8.6

Total Output Tokens

617,409

Total Input Tokens

321,297

Input Price

$0.035 / 1M

Output Price

$0.141 / 1M

Cache Read Price

$0.002 / 1M

Cache Write Price

N/A

Cache prices apply to input tokens. Reads reuse cached prompts; writes store them and can cost extra. Output tokens use the output price.

Tests Correct

Wrong Tests: 8

Attempt pass rate: 72.5%

Flaky tests

4

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

Response Time (avg)

49.04s

Response Time (max): 265.97s

Response Time (total): 1127.83s

Hamster playing table tennis

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

#218 Hy4 flash

medium
Cost
$0.002
Time
53.7s
Tokens
13,619 tok

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

Category Breakdown

Category Score Consistency Tests Correct
Agentic 4.7 3.1
Anti-AI Tricks 10.0 10.0
Coding 7.4 7.0
Combined 6.5 10.0
Data parsing and extraction 10.0 10.0
Domain specific 5.9 7.2
General Intelligence 10.0 10.0
Instructions following 10.0 10.0
Puzzle Solving 5.9 7.2
Tool Calling 10.0 10.0
Trivia 3.0 10.0

Compared models