Codesota · Models · RVT-2NVIDIA1 results · 1 benchmarks
Model card

RVT-2.

NVIDIAopen-sourceRobotic View Transformer 21 current SOTA

6x faster training than RVT, 19pp higher success rate on RLBench 18 tasks

§ 02 · Benchmarks

Every benchmark RVT-2 has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01RLBench—success-rate81.4%#1/32026-04-20source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 05 · Related models

Other NVIDIA models scored on Codesota.

Canary-Qwen-2.5B
2.5B params · 6 results
Canary-1B
1B params · 3 results
NV-Embed-v2
7B params · 3 results
Canary-1B-Flash
1B params · 2 results
Canary-180M-Flash
180M params · 1 result
MambaVision-L2
241M params · 1 result
NVIDIA-Nemotron-3-Super-120B-A12B-BF16
1 result
Nemotron 3 Nano (30B)
1 result
§ 06 · Sources & freshness

Where these numbers come from.

codesota-api
1
result
1 of 1 rows marked verified.