Event
Shanghai AI Lab releases decision model Intern-Decision and open-sources three weight tiers
1 reports · 1 sources
Shanghai Artificial Intelligence Laboratory has launched the multimodal decision model Intern-Decision, claiming its performance surpasses Jev and similar open-source models, and has released weights in three parameter tiers: 0.8B, 2B, and 4B. Across 7 benchmark evaluations including JevBench, the 4B model averaged a score of 90.02, ranking first among the evaluated models. On latency, in RTX 4090 tests the 0.8B and 2B models showed average end-to-end latency of about 33 milliseconds, roughly 30 decisions per second, while the 4B model was 44.16 milliseconds, faster than Jev's 109.70 milliseconds. The models will soon be integrated into the Intern-Duanyan scientific discovery platform for experimental interaction and automated scientific research in AI for Science scenarios.
Report timeline
- Shanghai AI Lab launches decision model Intern·Mingjue, outperforming Jev and releasing 0.8B, 2B, and 4B weights
Shanghai AI Laboratory has launched the decision model Intern·Mingjue (Decision), which outperforms Jev and is available in 0.8B, 2B, and 4B parameter versions. In RTX 4090 tests, the 0.8B and 2B models had an average end-to-end latency of about 33 milliseconds and about 30 decisions per second, while the 4B model was 44.16 milliseconds and Jev was 109.70 milliseconds.
Shanghai AI Laboratory - Embodied Intelligence ·
First-party reports
- Shanghai AI Lab launches decision model Intern·Mingjue, outperforming Jev and releasing 0.8B, 2B, and 4B weightsShanghai AI Laboratory - Embodied Intelligence