Hy4 flash

Hy4 flash is a reasoning model from Tencent in the Hy4 family. 8 benchmarks count toward its index, in 4 categories.

  • New
  • Reasoning
  • Input: text · image
  • Released 2026-11
RankedEnough results for an index.
IndexAll categories, out of 100.65.1±9.2Median 50.9
CoverageIndex weight that has results.65%
SpeedOutput tokens per second.72/sMedian 65/s
First tokenSeconds to the first output token.2.9 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.044Median $0.60
Output / 1MUS dollars per 1M output tokens.$0.176Median $2.50
ContextMaximum tokens in one request.1.05MMedian 262K
EloLMArena rating and rank.—Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
tencent/hy4-flash
Max outputMaximum output tokens per request.
66K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
<$0.01
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
2 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
OpenRouterRouter$0.035$0.1411.05M
Novita AIHost$0.07$0.2811M

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.5 counted71.122 / 221
CodingCode writing and repair.1 counted50.8—
ReasoningLogic problems and puzzles.None counted——
MultimodalTasks with images and text.1 counted53.6—
KnowledgeFacts and expert knowledge.1 counted79.1—
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.None counted——

Results

8 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Humanity's Last ExamKnowledge55.2%15 / 79
APEX-AgentsAgentic63.1%4 / 14
AutomationBenchAgentic52.2%10 / 33
DeepSWEAgentic64.6%36 / 61
OSWorld 2.1 (all 108 tasks)Agentic68.1%11 / 12
ZeroBench_main with PythonMultimodal45.8%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reusemodels.devMIT

Questions

How much does Hy4 flash cost?
$0.044 per 1M input tokens, $0.176 per 1M output tokens.
How good is Hy4 flash?
Its index is 65.1 out of 100. That is place 42 of 286 ranked models.
What is Hy4 flash best at?
Agentic: 71.1 out of 100, place 22 of 221.
How fast is Hy4 flash?
72 output tokens per second, 2.9 s to the first token.
What is the context window of Hy4 flash?
1.05M tokens.
When was Hy4 flash released?
November 2026.

Badge

Hy4 flash on Inference360: place 42 of 286, index 65.1
HTML
<a href="https://inference360.ai/models/hy4-flash"><img src="https://inference360.ai/badges/hy4-flash.svg" alt="Hy4 flash on Inference360: place 42 of 286, index 65.1" height="20"></a>
Markdown
[![Hy4 flash on Inference360: place 42 of 286, index 65.1](https://inference360.ai/badges/hy4-flash.svg)](https://inference360.ai/models/hy4-flash)

Compare with

Qwen3.8 Max65.1DeepSeek V4 Pro 081364.9GLM-5.365.6dots3-note Preview63.1Hy472.1

Lower price

dots3-note Preview63.1 · FreeGLM-5.3-Flash63.0 · $0.078/1MLing 3.1 Flash65.8 · $0.135/1M

More from Tencent

All models
Hy472.1Hy-Memory-30B-A3BUnrankedHy4 preview66.4Hy MT2 7BUnrankedAuKUnranked