Skip to content
Model Season every model has a season

Hy3

tencent/hy3-20260706
TencentChinaOpen weights

The 9th most-used model on the router over the last 7 days, with 2.7% of tokens. Weekly share peaked at 18.3% in the week of Jul 13, 2026, with 65T tokens processed since Jul 6, 2026.

Launch
Jul 6, 2026
Context
256k tokens
Modality
text → text
Reasoning
yes
License
Open weights
Input
$0.132 / 1M
Output
$0.528 / 1M
Share, last 7 days
2.7%
9th of 61 · Sep 4, 2026 to Sep 10, 2026
Peak weekly share
18.3%
week of Jul 13, 2026
Cumulative tokens
65T
since the week of Jul 6, 2026
Weeks in the top 10
9
of 9 weeks with volume

Weekly token share

Share of the router's weekly volume, since the first week with volume

0%5%10%15%20%25%Jul 6Jul 20Aug 3Aug 17Aug 31peak 18.3% on Jul 13, 2026
See the numbers
WeekShare
Aug 31, 20263.458%
Aug 24, 20265.891%
Aug 17, 20268.788%
Aug 10, 202613.211%
Aug 3, 202611.659%
Jul 27, 20268.496%
Jul 20, 20269.807%
Jul 13, 202618.346%
Jul 6, 202611.657%

Debuted in the week of Jul 6, 2026 and peaked at 18.3% 1 week later. The latest full week came in at 3.5%, 19% of peak.

What this model is used for new

Tasks where the model ranks among the leaders, and its share of each

Ranks among the leaders in 3 of the 29 classified tasks. Its largest share is in workflow execution, 4.3% of the task's tokens.

Snapshot of the rolling 7-day window through Sep 10, 2026. Share: the model's tokens in the task over the task's total tokens. Weight: the task's share of classified volume.

Where to run it

Endpoints in the source's catalog, from cheapest to most expensive on input

ProviderInputOutputContextQuantizationZero retention
DeepInfra$0.07$0.29256kfp8yes
Tencent$0.132$0.528256kfp8yes
Novita$0.14$0.58256knot reportedyes
GMICloud$0.14$0.58256kbf16no
Phala$0.15$0.64256knot reportedyes
AtlasCloud$0.20$0.80256kfp8no

6 endpoints across 6 providers, with input from $0.07 to $0.20 per 1M tokens. 4 offer zero data retention.

Prices in US dollars per million tokens, from the latest archived endpoint catalog.

Benchmarks

Artificial Analysis indexes and benchmarks run by the source, with cost per task

No Artificial Analysis index and no source-run benchmark for this model. A missing benchmark is not a low score.

Against the week's largest models

The five largest proprietary and five largest open-weights models over the last 7 days

ModelLab7-day shareBlended priceContextIntelligenceWeeks in top 10
Proprietary
GPT-5.6 LunaOpenAIOpenAI11.5%$0.451M37.56
Gemini 3.8 FlashGoogleGoogle2.1%$1.501M41.20
Muse Spark 1.3 ContributorMetaMeta1.5%$0.1251M0
Claude Opus 5AnthropicAnthropic1.4%$10.001M50.71
GPT-5.6 SolOpenAIOpenAI1.4%$4.001M47.10
Open weights
Hy4 previewTencentTencent15.5%$1.251M1
DeepSeek V4 Flash 0731DeepSeekDeepSeek10.0%$0.0941.31M34.55
GLM 5.3 FlashZ.ai (GLM)Z.ai (GLM)10.0%$0.2371.31M41.92
MiMo-V2.5XiaomiXiaomi4.2%$0.1751M22.315
DeepSeek V4 Flash 0423DeepSeekDeepSeek3.9%$0.1111M24.819
Hy3this pageTencentTencent2.7%$0.231256k9

Share over the last 7 days (Sep 4, 2026 to Sep 10, 2026). Blended price in US dollars per million tokens, weighted 75% input and 25% output. A dash means the number is not published, not zero.

What this page doesn't say. Share here is a slice of one router's traffic, where the deciding factor is usually price per token. It is not market, revenue or user share, and it does not tell you which model fits your workload.