by Baidu
A powerful multimodal Mixture-of-Experts chat model featuring 28B total parameters with 3B activated per token, delivering exceptional text and vision understanding through its innovative heterogeneous MoE structure with modality-isolated routing. Built with scaling-efficient infrastructure for high-throughput training and inference, the model leverages advanced post-training techniques including SFT, DPO, and UPO for optimized performance, while supporting an impressive 131K context length and RLVR alignment for superior cross-modal reasoning and generation capabilities.
| Signal | Strength | Weight | Impact |
|---|---|---|---|
| Recency2026-03-03T20:31:17.058Z | 96 | 15% | +14.4 |
| Capabilities2026-03-03T20:31:17.058Z | 57 | 25% | +14.3 |
| Context Window2026-03-03T20:31:17.058Z | 71 | 15% | +10.7 |
| Output Capacity2026-03-03T20:31:17.058Z | 65 | 10% | +6.5 |
| Versatility2026-03-03T20:31:17.058Z | 50 | 10% | +5.0 |
| Pricing Tier2026-03-03T20:31:17.058Z | 1 | 25% | +0.1 |
Cost Estimator
You save $38.13/month vs category average