Hy4 flash

Released
2026-11-16
Last refreshed
2026-11-16
Status
Researched today
ProprietaryCommercial use: restrictedMultimodalCodingRAGAgentsLong contextVisionJSON / Tool useCodingAgent
Tencent AI Lab releases · 13 in the last 12 months · this family litChangelog →
Specifications
Family
Hunyuan
Released
2026-11-16
Context
1M
Max output
64K
Parameters
224B (12B active)
Specialization
general
Openness
Proprietary
License
ProprietaryRestricted
Weights
Not released
Code
Unknown
Created by

AI innovations for societal improvement

Shenzhen, China
Founded 2016
Website
Pricing
Output / 1M
$0.281
Input / 1M
$0.070

Cheapest of 4 routes · Tencent Cloud TI Platform · cache read $0.003

About

Hy4 flash is Tencent Hy's lowest-cost model in the Hy4 family, released November 16, 2026. Tencent says it has 224B total parameters, including 64B Engram parameters, with 12B active per token, and combines hybrid attention, LatentMoE and Engram to shrink the KV cache to 1/6 of Hy4's. Like Hy4 it is natively multimodal, taking text and image input with a 1M-token context window, and supports reasoning, function calling, structured outputs and prompt caching. Tencent positions it for high-volume everyday work such as coding assistance, agentic workflows and document and chart understanding. API-only; weights are not released.

Providers(4)

Compare all 4

Compare API pricing across 4 tracked provider routes for input and output tokens, batch, and cached reads when available.

ProviderInput / 1MOutput / 1MCacheRoute
Tencent Cloud TI Platform$0.070$0.281read $0.003
Serverless
Novita AI$0.070$0.281-
Serverless
OpenRouter---
ServerlessPartial
Vercel AI Gateway---
ServerlessPartial

Capabilities

VisionMultimodalReasoningJSON / Tool useStructured OutputsPrompt Caching

Benchmark peer barsfor Coding

No task-mapped benchmark peers are available for this model yet.

Migration checks

No linked migration route is available for this model yet.

API versions

hy4-flashtencent/hy4-flash