Models · VLA (vision-language-action)

smolvla/so101

checking availability…

A specific checkpoint, edge-inference/smolvla-so101-pick-orange, trained in simulation, for the SO-101 arm. It answers one observation with the next 50 actions. It is not a general-purpose controller; the task sentence is passed to the model, but only the task it was trained on is expected to work.

Price

per call: $0.005 · 50 tokens
per action: $0.0001
billed: every returned action, whether or not your robot plays it

What you send

observation.state [6]: shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper · normalized: arm -100..100, gripper 0..100
observation.images.front [480, 640, 3]: front camera, RGB uint8 (JPEG on the wire)
observation.images.wrist [480, 640, 3]: wrist camera, RGB uint8 (JPEG on the wire)
task text: passed to the policy

What comes back

50 actions × 6 joints, to be played at 30 Hz — 1.67 s of motion per call.

Units before motion.Values are LeRobot normalized (arm -100..100, gripper 0..100), not degrees. The semantics label depends on where you read it — the model catalog still answers so101-degrees-gripper-percent-v1, a lease and the wire answer the corrected label — and the data is normalized either way, so never branch on it (units, explained). LeRobot’s so101_follower defaults to degrees: run it with --robot.use_degrees=false. What the arm does if the connection drops.

Measured

policy: 367–417 ms
on the runner: 382–446 ms
client round trip: 552–1141 ms

From a recorded probe on 2026-10-25 06:00Z. The round trip includes that client’s network; yours will differ. What each number measures.

Try it

One billed call with synthetic inputs ($0.005). No robot moves.

Code

pip install agi-diy
agi login
agi run smolvla/so101 --probe -n 1

Needs Python 3.12 — LeRobot is installed on first use, requires 3.12 or later and does not run on 3.14 yet.

Or run a test call from the dashboard, or drive the simulated SO-101.

Licences and data

Checkpoint	edge-inference/smolvla-so101-pick-orange @ `71cf4a9` · 1.2 GB	apache-2.0
Base model	lerobot/smolvla_base	apache-2.0
Training data	LightwheelAI/leisaac-pick-orange	apache-2.0

Licences as declared on the Hub at the pinned revision. Policies, explained.