Advertise here
#304

Hy4 preview

Tencent Release: 2026-08-29 Tested on: 2026-10-01 04:19 tencent/hy4-preview::low
770B total (49B active)MoEOpen source
(low) (none)

Summary

Hy4 preview scores 5.4 on AI BENCHY and ranks #304. It has 6.5 reliability, a 21.7% pass rate, $1.844 total cost, and 207.21s average response time.

What makes Hy4 preview unique: It stands out most in General Intelligence, where it ranks #2, while Tool Calling is its weakest area at #12. It uses unusually many reasoning tokens, which can help explain its slower or more expensive runs.

Model facts

Researched on 2026-08-29

Reported
Parameters
770B total (49B active)
Architecture
MoE
Availability
Open source
License
Apache-2.0

Vendor-reported backbone counts exclude the native 10B total / 0.7B active MTP layer used for speculative decoding.

Parameter source 1 Parameter source 2 Availability source

Score

5.4

Consistency

8.2

Reliability

6.5

Total Cost (Current Price)

$1.844

Total Output Tokens

687,477

Total Input Tokens

162,180

Input Price

$0.751 / 1M

Output Price

$2.251 / 1M

Cache Read Price

$0.038 / 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: 20

Attempt pass rate: 21.7%

Flaky tests

4

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

Response Time (avg)

207.21s

Response Time (max): 597.45s

Response Time (total): 4765.91s

Wrong answer: 13 API error: 5 Did not follow instructions: 1 No answer: 1

Hamster playing table tennis

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

#304 Hy4 preview

low
Provider returned error
Cost
$0.000
Time
1.1s
Tokens
0 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-10-01 04:19 Re-test 5.4 6.5 $1.844 Current run
2026-08-29 14:28 Initial run 4.3 1.6 $1.433 Compare

Run comparison

RunBenchmark coverageScoreConsistencyReliabilityTests CorrectFlaky testsTotal Output TokensTotal Input TokensTotal CostResponse Time (avg)
2026-10-01 04:19 · Current run69/69 attempts5.48.26.53/234687,477162,180$1.844207.21s
2026-08-29 14:28 · Initial run66/66 attempts4.35.01.63/2213559,13141,319$1.433189.47s
Difference—+1.0+3.2+4.90-9+128346+120861+$0.411+17745ms

Benchmark coverage differs: 69/69 attempts (Target: 3 repeats per test) versus 66/66 attempts (Target: 3 repeats per test). Totals and repeat-sensitive metrics are not directly comparable.

Price History

Historical pricing data for this model from OpenRouter.

Date Input Price Output Price Cache Read Price Cache Write Price
2026-10-06 16:43 $0.751 / 1M $2.251 / 1M $0.038 / 1M N/A

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

Hy4 previewlowvsNorth Mini CodemediumFree AvailableHy4 previewlowvsMiMo-V2.6-FlashnoneHy4 previewlowvsQwen3.8 27BnoneFree AvailableHy4 previewlowvsStep 3.5 FlashmediumHy4 previewlowvsSeed-2.0-CodenoneHy4 previewlowvsEmber-1noneHy4 previewlowvsGemma 4 26B A4BnoneFree AvailableHy4 previewlowvsGPT-5.6 LunanoneHy4 previewlowvsGPT-6 LunanoneHy4 previewlowvsGranite 4.2 8Blow

Category Breakdown

Category Score Consistency Tests Correct
Agentic 4.7 1.6
Anti-AI Tricks 6.0 8.9
Coding 5.2 8.6
Combined 2.9 5.8
Data parsing and extraction 8.7 10.0
Domain specific 4.5 10.0
General Intelligence 7.5 10.0
Instructions following 9.8 10.0
Puzzle Solving 5.3 5.0
Tool Calling 3.0 10.0
Trivia 3.0 10.0

Compared models