AI Model Radar
中文 EN

Hunyuan HY3-FP8

Tencent / Hunyuan
Model overview

Hy3-FP8

Tencent's FP8-quantized Hy3 instruct checkpoint, tracked as a separate open-weight model surface for local inference, coding, and agent w...

general LLMreasoningcodinglong contextagentchatcode generationsoftware engineering agentlocal inference
CompareCompanyRankingsPublic signals
Representative version

Hy3-FP8

CompanyTencent / Hunyuan
Release2026-07-04
Release typeModel checkpoint release
Parameters295B / 21B active
ArchitectureFP8-quantized mixture-of-experts transformer
Context262,144
Input modalitiestext / code
Output modalitiestext / code / tool_call
Tool use / function callingConfirmed: Tool use, Function calling
Structured output / reasoningConfirmed: Reasoning
Vision / audio / videoNot disclosed
Open weights / LicenseWeights available / Apache-2.0
License sourcetencent-hy3-fp8-current · 2026-07-15
Vendor APINot disclosed
API source pending verification
OpenRouterNot separately listed on OpenRouter
PricingNo reliable public source
BenchmarkNo reliable public source
Public signals1 runtime / access record / 1 activity snapshot
Use casescomplex reasoning / coding and review / Agent workflows / local inference
Claim evidenceRelease / model specs / weightsTencent Hy3-FP8 model cardHy3-FP8 · 2026-07-04
OpenRouter

Pricing / context / availability

OpenRouterNot separately listed on OpenRouter
OpenRouter listingNot separately listed in the current OpenRouter public list
Mapped versionHy3-FP8
tencent-hy3-fp8-current
PricingNo reliable public source
Context262144
Speed/LatencyNo reliable public source
Snapshot time2026-07-15T20:11:41.551Z
SourceView source
Public activity counters

Cumulative Hugging Face model-card downloads / likes / spaces

PlatformEntryDownloadsLikesSpacesTimeSource
huggingfacetencent/Hy3-FP812,1065802026-07-15T20:11:41.551ZView source
Operations

Runtime / access signals

TimeSignalOriginal textSource
07/07/2026, 09:06 PM UTCvllm 0.24 RuntimeError: HYV3ReasoningParser reasoning parser could not locate think start/end tokens in the...(APIServer pid=143) Traceback (most recent call last): (APIServer pid=143) File "/root/vllm/.venv/bin/vllm", line 10, in <module (APIServer pi...Source
Evidence gaps

Public display boundary

OpenRouter not separately verifiedNo reliable benchmark sourceNo linkable real reviewSpeed/latency not disclosed