SenseTime · robot AI model
SenseNova-RoboRSI-27B-Preview
China · 2026
Sources retrieved: 12 November 2026 (A, B)
- Model type
- Embodied reasoningSource and note: Model typeA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Developer wording: “SenseNova-RoboRSI-27B-Preview is an open-weight 27B vision-language model for embodied agents.” “The model is derived from Qwen3.8-27B, whose weights were further post-trained by SenseNova for embodied reasoning and interaction.”
- Inputs
- Camera LanguageSource and note: InputsA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “SenseNova-RoboRSI-27B-Preview is an open-weight 27B vision-language model for embodied agents.” “The model actively inspects visual details, reasons about spatial relationships, and corrects actions based on execution feedback during complex manipulation tasks.”
- Access
- Open weights (download)Source and note: AccessA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “Download the full checkpoint, including the tokenizer, processor, generation configuration, and chat template:”
- Parameters
- 27BSource and note: ParametersA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “SenseNova-RoboRSI-27B-Preview is an open-weight 27B vision-language model for embodied agents.”
- Base model
- Qwen3.8-27BSource and note: Base modelA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “The model is derived from Qwen3.8-27B, whose weights were further post-trained by SenseNova for embodied reasoning and interaction.”
- Licence type
- Free, commercial useSource and note: Licence typeA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Developer wording: “The model weights and model card are released under the Apache License 2.0.”
No photograph
Specifications
7 of 14 stated 7 not stated
4 fields hidden 3 not-stated fields with a note stay ·
How to read the figuresShow guideClose
- not statedNo source
- 0Reported as zero
- ASource
Select a source marker for source details and notes.
Release0/1 disclosed
ReleasedNot stated
Source and note: ReleasedB huggingface.co/api/models/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Hugging Face repository metadata give the repository creation time as “2026-10-16T09:26:09.000Z”.
Input and output2/4 disclosed
Model typeEmbodied reasoning
Source and note: Model typeA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Developer wording: “SenseNova-RoboRSI-27B-Preview is an open-weight 27B vision-language model for embodied agents.” “The model is derived from Qwen3.8-27B, whose weights were further post-trained by SenseNova for embodied reasoning and interaction.”
InputsCamera Language
Source and note: InputsA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “SenseNova-RoboRSI-27B-Preview is an open-weight 27B vision-language model for embodied agents.” “The model actively inspects visual details, reasons about spatial relationships, and corrects actions based on execution feedback during complex manipulation tasks.”
What it controlsNot stated
Source and note: What it controlsA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: the model learns from interaction experience, “developing capabilities for planning, action decisions, and feedback-based correction in simulation and on real robots”.
Degrees of freedom controlledNot stated
Size and speed1/2 disclosed
Parameters27B
Source and note: ParametersA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “SenseNova-RoboRSI-27B-Preview is an open-weight 27B vision-language model for embodied agents.”
Control rateNot stated
Where it runs and what it is built on1/2 disclosed
On robot or cloudNot stated
Base modelQwen3.8-27B
Source and note: Base modelA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “The model is derived from Qwen3.8-27B, whose weights were further post-trained by SenseNova for embodied reasoning and interaction.”
Openness3/3 disclosed
AccessOpen weights (download)
Source and note: AccessA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “Download the full checkpoint, including the tokenizer, processor, generation configuration, and chat template:”
Licence documentApache-2.0
Source and note: Licence documentA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Model card: “The model weights and model card are released under the Apache License 2.0.”
Licence typeFree, commercial use
Source and note: Licence typeA huggingface.co/sensenova/SenseNova-RoboRSI-27B-Preview · Retrieved 12 November 2026Note. Developer wording: “The model weights and model card are released under the Apache License 2.0.”
The developer’s own claims0/2 disclosed
Learns new tasksNot stated
Works across robotsNot stated
Source and note: Works across robotsYes 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
SenseTime's SenseNova website states: “SenseTime explores embodied intelligence and open-sources the SenseNova-RoboRSI framework and a 27B embodied model.”
SenseTime's website gives its address as “SenseTime Building, No. 1900 Hongmei Road, Xuhui District, Shanghai”.
Sources
One line per source, not per figure: a source has one retrieval date.
- Ahuggingface.co/sensenova/SenseNova-RoboRSI-27B-PreviewMain source. Retrieved 12 November 2026
- Bhuggingface.co/api/models/sensenova/SenseNova-RoboRSI-27B-PreviewRetrieved 12 November 2026
