Hugging Face LeRobot · SmolVLA · active

SmolVLA 450M

Hugging Face LeRobot's compact open vision-language-action model designed for consumer hardware and affordable robot arms.

lerobot/smolvla_base
Statusactive
InputText, Image, Robot state
OutputRobot action
Context
Total / active parameters450M /
Licenseapache-2.0
APINo
Open weights / self-hostYes / Yes

This page represents one immutable developer model ID. Serving endpoints, pricing, regions, quantizations, and service tiers attach separately and never create duplicate model records.

Provider availability

Serving endpoints and pricing

No directly sourced provider offerings yetPricing and availability stay unknown until collected from each serving provider.

Technical specification

Developer model IDlerobot/smolvla_base
Version450M
ReleasedJun 3, 2025
ArchitectureSmolVLM2 backbone with flow-matching action expert
Maximum outputUnknown
Knowledge cutoffNot reported

Capabilities

asynchronous-inferencefine-tuninglow-cost-hardwaremanipulation
ReasoningNo
VisionYes
Tool callingNo

Version lineage

PredecessorUnknown
SuccessorsUnknown
Aliaseslerobot/smolvla_base

Explicit unknowns

  • Maximum output
  • Tokenizer
  • Knowledge cutoff
  • Provider pricing

Comparable peers

Related model comparisons

All comparisons →

Questions

SmolVLA 450M FAQs

What is SmolVLA 450M?+

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.

What inputs and outputs does SmolVLA 450M support?+

SmolVLA 450M's current record lists Text, Image, and Robot state as input and Robot action as output.

Is SmolVLA 450M available through an API?+

SmolVLA 450M is explicitly marked as not API-available in the current record.

How large is SmolVLA 450M?+

The current record lists 450M total parameters and an unknown active parameter count. Its architecture is SmolVLM2 backbone with flow-matching action expert.

What capabilities and tasks does SmolVLA 450M support?+

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.

Can SmolVLA 450M be self-hosted?+

SmolVLA 450M is an open-weight model. Self-hosting is supported, and the recorded license is apache-2.0.

Does SmolVLA 450M have related versions or aliases?+

No predecessor is recorded. No successor is recorded. Recorded aliases are lerobot/smolvla_base.

Which models can SmolVLA 450M be compared with?+

The current compatible peer set includes OpenVLA 7B, GR00T N1.7 3B, MolmoAct 2, Helix 02, and pi 0.7. Compatibility requires shared sourced modalities and tasks; it does not imply that one model is better. Browse model comparisons

How current is the Model Markets record for SmolVLA 450M?+

SmolVLA 450M was last verified Aug 28, 2026 from Hugging Face's primary source, “SmolVLA: Efficient Vision-Language-Action Model trained on LeRobot Community Data.” The evidence is classified as official fact.

Primary evidence

Sources and observation date