Identity
- Status
- active
- Developer ID
lerobot/smolvla_base- Released
- Jun 3, 2025
- Version
- 450M
- License
- apache-2.0
Hugging Face LeRobot's compact open vision-language-action model designed for consumer hardware and affordable robot arms.
0 current hosted offerings
This record is downloadable or self-hosted. Its cost depends on the infrastructure and serving configuration, not a universal model price.
At a Glance
lerobot/smolvla_basePublisher Media
Embodied Intelligence
Compatibly licensed LeRobot community datasets totaling fewer than 30,000 episodes in the cited release.
Verified Platform Links
| Robot | Relationship | Form | Availability | Evidence note | Source |
|---|---|---|---|---|---|
SO-101Hugging Face LeRobot | Evaluation embodiment | Fixed manipulator | Open build | Hugging Face reports real-world SO-101 evaluation and deployment support for SmolVLA. | Official fact |
This page represents one developer model product. Dated API snapshots, serving endpoints, pricing, regions, quantizations, and service tiers attach as versioned aliases or provider details and never create duplicate public model records.
Primary Evidence
Comparable Peers
| A | Pair | B | Context |
|---|---|---|---|
SmolVLA 450MHugging Face LeRobot | vs | MolmoAct 2Ai2 | cross-developer peersrobotics |
SmolVLA 450MHugging Face LeRobot | vs | Helix 02Figure | cross-developer peersrobotics |
SmolVLA 450MHugging Face LeRobot | vs | Gemini Robotics 2Google DeepMind | cross-developer peersrobotics |
SmolVLA 450MHugging Face LeRobot | vs | GR00T N1.7 3BNVIDIA | cross-developer peersrobotics |
SmolVLA 450MHugging Face LeRobot | vs | OpenVLA 7BOpenVLA Research Team | cross-developer peersrobotics |
SmolVLA 450MHugging Face LeRobot | vs | pi 0.7Physical Intelligence | cross-developer peersrobotics |
Questions
Hugging Face LeRobot's compact open vision-language-action model designed for consumer hardware and affordable robot arms. It is developed by Hugging Face LeRobot and its current sourced lifecycle status is active.
SmolVLA 450M's current record lists Text, Image, and Robot state as input and Robot action as output.
SmolVLA 450M is classified as a vision-language-action model. Its action representation is Continuous action chunks from a flow-matching action expert, its control architecture is SmolVLM2 backbone with flow-matching action expert, and inference is recorded as on device.
The developer reports SO-100, SO-101, LeKiwi, and LIBERO Franka as supported or represented embodiments. Model Markets also has sourced model-to-platform links for SO-101; each link states whether it is native, demonstrated, used for training, or used for evaluation.
The current record lists 450M total parameters and an unknown active parameter count. Its architecture is SmolVLM2 backbone with flow-matching action expert.
SmolVLA 450M's recorded capabilities are asynchronous-inference, fine-tuning, low-cost-hardware, and manipulation. Its supported tasks are robot-control, dexterous-manipulation, and pick-and-place.
Compatibly licensed LeRobot community datasets totaling fewer than 30,000 episodes in the cited release.
SmolVLA 450M is explicitly marked as not API-available in the current record.
SmolVLA 450M has no verified hosted API offering, so no provider token price applies. It is a downloadable or self-hosted artifact whose cost depends on the serving infrastructure and workload.
SmolVLA 450M is an open-weight model. Self-hosting is supported, and the recorded license is apache-2.0.