Tencent Robotics XChina2026

Hy-Embodied-VLM-1.0

An AI model: a vision-language-action model (VLA) takes camera images and an instruction and turns them into movements for a robot. This page describes the software, not a robot.

Robotsnot stated

Key figures as stated by the developer

4 of 6 stated · 2 gaps

Model type not stated
Controls not stated
Inputs ⁠A camera language
Access ⁠A open download
Parameters ⁠A approximately 30,000,000,000
Base model ⁠A Hy3-A3B language backbone with Hy-ViT2 vision encoder

How to read the figures

14kgA30.9lbA: The letter in the grey square after a figure points to the source the figure came from. Robots with several sources carry several letters.
74.5lb: Imperial figures are our own conversion unless the manufacturer states an imperial figure itself. The source letter always belongs to the metric figure — a conversion has no source of its own.

The four states

Versions and notes in the key figures (6)

Model type

Note The developer calls it something else and does not label it vision-language-action, world action model or embodied reasoning: "An efficient Mixture-of-Experts vision–language foundation model for embodied agents in the physical world".

Controls

Note The model gives text answers, plans and navigation decisions, not robot motion: "Hy-Embodied-VLM-1.0 powers closed-loop navigation with a single model for language grounding, spatial memory, planning, and goal verification."

Inputs

The model card lists the vision inputs as "Image (up to 128 per prompt)".

Access

"Weights are available on" Hugging Face.

Parameters

Tencent gives an approximate total, "~30B total". "Only ~3B parameters are activated per token".

Base model

Tencent's description: "Combines the Hy3-A3B language backbone with the Hy-ViT2 vision encoder".

Full specifications

5 of 12 fields stated

The letter points to the source at the foot of the page. Fields without a figure are dashed: the developer does not state them — that is not the same as zero, and not the same as no.
Field	Value	Source and caveats
Input and output
Model type	not stated	Note The developer calls it something else and does not label it vision-language-action, world action model or embodied reasoning: "An efficient Mixture-of-Experts vision–language foundation model for embodied agents in the physical world". ⁠A
Inputs	`camera` `language`	A The model card lists the vision inputs as "Image (up to 128 per prompt)".
Controls	not stated	Note The model gives text answers, plans and navigation decisions, not robot motion: "Hy-Embodied-VLM-1.0 powers closed-loop navigation with a single model for language grounding, spatial memory, planning, and goal verification." ⁠A
Degrees of freedom controlled	not stated
Size and speed
Parameters	approximately 30,000,000,000	A Tencent gives an approximate total, "~30B total". "Only ~3B parameters are activated per token".
Control rate	not stated
Where it runs and what it is built on
Runs where	not stated	Note "The full BF16 model requires ~86 GB across GPUs; a single 8×80 GB node is sufficient." Tencent does not say that the model runs on the robot. ⁠A
Base model	Hy3-A3B language backbone with Hy-ViT2 vision encoder	A Tencent's description: "Combines the Hy3-A3B language backbone with the Hy-ViT2 vision encoder".
Openness
Access	open download	A "Weights are available on" Hugging Face.
License	Apache-2.0	B "Hy-Embodied-VLM-1.0 is licensed under the Apache-2.0." "refers to the training code, the inference code, parameters, and weights".
The developer’s own claims
Learns new tasks	not stated	Note Fine-tuning code is on Tencent's list of planned releases, not yet released. ⁠A
Works across robots	not stated	Yes only if the developer itself says the model runs on different robot models, for example robots from two manufacturers or two of one manufacturer’s models. Several units of the same robot do not count.

Notes

Tencent describes Hy-Embodied-VLM-1.0 as an embodied foundation model for agents in the physical world; it was released on 15 July 2026.
Tencent says the model powers closed-loop navigation with language grounding, spatial memory, planning and goal verification.
The model is a Mixture-of-Experts vision-language model built by Tencent Robotics X, the Hy Vision Team and Futian Laboratory.

Sources

A Manufacturer product page. github.com/Tencent-Hunyuan/HY-Embodied · Retrieved 2026-10-13
B huggingface.co/tencent/Hy-Embodied-VLM-1.0/blob/main/LICENSE · Retrieved 2026-10-13

One line per source, not per figure: a source has one retrieval date. The letter beside a figure points to the line it came from.