OpenBMB · robot AI model
MiniCPM-RobotManip
China · 2026
Sources retrieved: 9 November 2026 (A, B)
- Model type
- Vision-language-action (VLA)Source and note: Model typeB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. Hugging Face model card: “MiniCPM-RobotManip is a 1.5B vision-language-action model for embodied manipulation”.
- What it controls
- Arms and handsSource and note: What it controlsA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. Normalised from OpenBMB wording: the model is built for manipulation (“MiniCPM-RobotManip is a 1.5B vision-language-action model for embodied manipulation”) and “The model returns an action chunk of shape (30, 80)”.
- Inputs
- Camera Language Robot stateSource and note: InputsA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. README, inference: “Provide at least one image and a language instruction; robot state defaults to zeros if omitted.”
- Access
- Open weights (download)Source and note: AccessB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. Hugging Face model card: “Model weights and code are open-sourced under the” Apache-2.0. The repository is listed as not gated.
- Parameters
- 1.5BSource and note: ParametersB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. Hugging Face model card: “MiniCPM-RobotManip is a 1.5B vision-language-action model for embodied manipulation”. README model zoo: “A 1.5B vision-language-action model for Robot Manipulation”.
- Licence type
- Free, commercial useSource and note: Licence typeB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. “Model weights and code are open-sourced under the” Apache-2.0.
Specifications
8 of 14 stated 6 not stated
2 fields hidden 4 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.
Release1/1 disclosed
Released19 July 2026
Source and note: ReleasedA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. [A] README news “[2026.07.19]”: “We release and open-source MiniCPM-Robot”.
Input and output3/4 disclosed
Model typeVision-language-action (VLA)
Source and note: Model typeB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. Hugging Face model card: “MiniCPM-RobotManip is a 1.5B vision-language-action model for embodied manipulation”.
InputsCamera Language Robot state
Source and note: InputsA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. README, inference: “Provide at least one image and a language instruction; robot state defaults to zeros if omitted.”
What it controlsArms and hands
Source and note: What it controlsA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. Normalised from OpenBMB wording: the model is built for manipulation (“MiniCPM-RobotManip is a 1.5B vision-language-action model for embodied manipulation”) and “The model returns an action chunk of shape (30, 80)”.
Degrees of freedom controlledNot stated
Source and note: Degrees of freedom controlledA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. OpenBMB gives the shape of the action chunk, not the number of robot joints it controls: “The model returns an action chunk of shape (30, 80)”.
Size and speed1/2 disclosed
Parameters1.5B
Source and note: ParametersB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. Hugging Face model card: “MiniCPM-RobotManip is a 1.5B vision-language-action model for embodied manipulation”. README model zoo: “A 1.5B vision-language-action model for Robot Manipulation”.
Control rateNot stated
Source and note: Control rateA github.com/OpenBMB/MiniCPM-Robot · Retrieved 9 November 2026Note. OpenBMB gives inference speeds, not a control rate: “With H100, BF16, and single-frame input, model-forward latency per decision step is 120 ms”. README news, third-party inference engine: PhyAI “adds Day-0 support for MiniCPM-Robot, increasing inference throughput on NVIDIA H20 from 10 Hz to 37 Hz”.
Where it runs and what it is built on0/2 disclosed
On robot or cloudNot stated
Base modelNot stated
Source and note: Base modelB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. OpenBMB names the visual token compression it inherits, not a base model: “Inherits MiniCPM-V 4.6's visual token compression”.
Openness3/3 disclosed
AccessOpen weights (download)
Source and note: AccessB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. Hugging Face model card: “Model weights and code are open-sourced under the” Apache-2.0. The repository is listed as not gated.
Licence documentApache-2.0
Source and note: Licence documentB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. “Model weights and code are open-sourced under the” Apache-2.0. Tag: “license: apache-2.0”.
Licence typeFree, commercial use
Source and note: Licence typeB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. “Model weights and code are open-sourced under the” Apache-2.0.
The developer’s own claims0/2 disclosed
Learns new tasksNot stated
Works across robotsNot stated
Source and note: Works across robotsB huggingface.co/openbmb/MiniCPM-RobotManip · Retrieved 9 November 2026Note. OpenBMB describes one set of weights across tasks, not across robot models: the policy “uses one set of weights across all downstream tasks”.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
OpenBMB describes MiniCPM-RobotManip on Hugging Face as a 1.5B vision-language-action model for embodied manipulation.
MiniCPM-RobotManip is one of the two first models of the MiniCPM-Robot family; the other is MiniCPM-RobotTrack.
Sources
One line per source, not per figure: a source has one retrieval date.
- Agithub.com/OpenBMB/MiniCPM-RobotRetrieved 9 November 2026
- Bhuggingface.co/openbmb/MiniCPM-RobotManipMain source. Retrieved 9 November 2026
