Hy3 model profile

Evidence snapshot: 2026-09-06

Hy3 is a Tencent model profile with a 262,144-token context window. The dated catalog snapshot records text modalities and interfaces for prompt caching, reasoning controls, structured output, text generation, tool calling.

Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort: a direct no-think mode by default, plus low and high chain-of-thought modes for complex math, coding, and multi-step problems. With a 256K context window, Hy3 targets long-horizon tasks, including improved coreference resolution, multi-turn constraint tracking, and stable tool-calling that generalizes across agent scaffoldings. Tencent positions it as a reliable, cost-effective option across coding, document processing, financial analysis, game development, and frontend design, with a strong emphasis on grounded, anti-hallucination behavior that answers when grounded and flags when evidence is missing rather than fabricating.

Hy3 is a reference profile: Tencent publishes it, Kendr documents it for comparison, and it carries no Kendr alias to call.

Hy3 is an independent knowledge profile. It does not claim that Kendr hosts the model, quote a Kendr customer price, promise API availability, or advertise a Kendr routing receipt.

Hy3 accepts text and returns text. The snapshot records prompt caching, reasoning controls, structured output, text generation, and tool calling as capabilities or interfaces.

The dated reference snapshot lists $0.0825 input and $0.33 output per million tokens. Cached input is $0.020625 per million tokens.

Hy3 ranked #7 in the cited trailing-7-day third-party catalog usage dataset as of 2026-09-06. Observed within the cited trailing-7-day third-party catalog usage dataset; this is not global AI market share.

Profile data was reviewed for the 2026-09-06 snapshot. Claims retain their source dates, and unavailable fields are shown as unavailable instead of estimated.

Profile facts

Profile type: Model research profile
Provider or publisher: Tencent
Context window: 262,144 tokens
Snapshot date: 2026-09-06
Knowledge cutoff: Not disclosed
Reference catalog ID: tencent/hy3

When to pick Hy3

Hy3 is a reference profile: Tencent publishes it, Kendr documents it for comparison, and it carries no Kendr alias to call.

Each condition below is derived from this model's own values in the 2026-09-06 snapshot, compared only against models publishing the same field.

Reach for it when the model has to call tools. Tool calling is recorded in the snapshot, so this route can drive an agent loop rather than only answer in prose.
Reach for it when you want to trade depth against cost per call. Reasoning effort is selectable across high, low, none, so one route can serve both cheap and deep work.
Look elsewhere when you need to call it through Kendr today. This is a research reference profile with no Kendr alias. The facts and evidence here are published for comparison; the model is not routable on Kendr.

Context, modalities, and identity

Hy3 accepts text and returns text. The snapshot records prompt caching, reasoning controls, structured output, text generation, and tool calling as capabilities or interfaces.

Provider or publisher: Tencent
Context window: 262,144 tokens
Maximum output: 128,000 tokens
Input modalities: text
Output modalities: text
Knowledge cutoff: Not disclosed

Model overview

Reference model ID: tencent/hy3
Canonical version: tencent/hy3-20260706
Hugging Face ID: tencent/Hy3

Capabilities and supported controls

Capabilities and parameters are reported from the dated catalog and source records; their presence does not guarantee identical behavior across every provider route.

Supported parameters: frequency_penalty, include_reasoning, logit_bias, max_completion_tokens, max_tokens, min_p, presence_penalty, reasoning, reasoning_effort, repetition_penalty, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p

prompt caching
reasoning controls
structured output
text generation
tool calling

Dated pricing snapshot

The dated reference snapshot lists $0.0825 input and $0.33 output per million tokens. Cached input is $0.020625 per million tokens.

These are dated third-party reference prices, not Kendr prices or an availability offer. Routes, tiers, caching, region, and provider policy can change the landed price.

Price date: 2026-09-06
Input: $0.0825 per 1M tokens
Cached input: $0.020625 per 1M tokens
Output: $0.33 per 1M tokens

Provider routes, performance, and uptime

The snapshot retains 6 provider endpoints. Provider prices, context limits, p50 performance, and uptime can differ by route and are not Kendr guarantees.

Active provider endpoints: 5
Best p50 latency: 0.77 s
Best p50 throughput: 105 tok/s
Availability with routing: 99.68% over the sampled window
Availability without routing: 98.88% over the sampled window
Performance date: 2026-09-06

Provider	Quantization	Input / 1M	Output / 1M	Cache read / 1M	Context	p50 latency	p50 throughput	Uptime (1d)
Tencent Cloud	fp8	$0.0825 / 1M	$0.33 / 1M	$0.020625 / 1M	262,144 tokens	2.14 s	81.5 tok/s	99.90%
NovitaAI	unknown	$0.14 / 1M	$0.58 / 1M	$0.035 / 1M	262,144 tokens	2.46 s	76 tok/s	99.83%
DeepInfra	fp8	$0.14 / 1M	$0.58 / 1M	$0.035 / 1M	262,144 tokens	0.77 s	55 tok/s	99.61%
Phala	unknown	$0.15 / 1M	$0.64 / 1M	$0.04 / 1M	262,144 tokens	1.35 s	18 tok/s	99.92%
AtlasCloud	fp8	$0.2 / 1M	$0.8 / 1M	$0.05 / 1M	262,144 tokens	1.19 s	105 tok/s	100.00%
GMICloud	bf16	$0.126 / 1M	$0.522 / 1M	$0.0315 / 1M	262,144 tokens	2.23 s	19 tok/s	99.69%

Benchmark evidence and limitations

Third-party catalog fields normalized from the cited model dataset; benchmark methodology and coverage differ by source.

Design Arena category	Elo	Win rate	Rank
3d	1214	43.8%	#44
codecategories	1193	41.1%	#55
dataviz	1145	36.1%	#76
gamedev	1166	38.6%	#64
uicomponent	1185	40.2%	#60
website	1195	41.3%	#60

Popularity and market context

Third-party catalog rank: #7
Global market share: Not inferred
Observation date: 2026-09-06

Frequently asked questions

When should I use Hy3?

Hy3 is a reference profile: Tencent publishes it, Kendr documents it for comparison, and it carries no Kendr alias to call. Reach for it when the model has to call tools: Tool calling is recorded in the snapshot, so this route can drive an agent loop rather than only answer in prose. Reach for it when you want to trade depth against cost per call: Reasoning effort is selectable across high, low, none, so one route can serve both cheap and deep work. Look elsewhere when you need to call it through Kendr today: This is a research reference profile with no Kendr alias. The facts and evidence here are published for comparison; the model is not routable on Kendr.

What is Hy3?

What context window does Hy3 have?

The dated profile lists 262,144 tokens of context and up to 128,000 output tokens. Provider routes, variants, and runtime configuration can impose lower effective limits.

What benchmark evidence is available for Hy3?

The 2026-09-06 snapshot includes 6 Design Arena categories. Publisher-reported launch results are labeled and are not presented as Kendr measurements; scores from different suites are not treated as interchangeable. Third-party catalog fields normalized from the cited model dataset; benchmark methodology and coverage differ by source.

Is Hy3 available through Kendr?

This is an independent knowledge profile, not a Kendr-hosted availability claim. Check Kendr's live public model API for currently enabled Kendr aliases.

Does the popularity rank represent Hy3's global market share?

Sources and evidence dates

Hy3 third-party catalog record (catalog, checked 2026-09-06)
Third-party model catalog methodology (methodology, checked 2026-09-06)
Design Arena leaderboard and methodology (benchmark, checked 2026-09-06)