WeMM-Video-35B-A3B vs Inkling 2 Mini
Compare documented capabilities and access. A dash means a comparable figure is unavailable.
| Specification | ||
|---|---|---|
| Task | Audio-visual understanding | Multimodal reasoning and coding |
| Published capabilities | Answers questions in Chinese or English about video frames and their soundtrack. Supports direct responses and a thinking mode. | Reuses its middle layers three times per token. Thinking Machines reports 48.2% on DeepSWE v1.1. |
| Model size | 35B total / 3B active MoE backbone | 26.4B total / 3.4B active parameters |
| Access | Public BF16 checkpoint and quantized releases | Downloadable weights and Tinker inference and customization |
| Result | Text responses, including descriptions and speech transcripts | — |
| Operating limits | — | Native mixed-precision checkpoint occupies approximately 20 GB. |
| Sources | Model weights · Additional documentation | Model weights · Additional documentation |
| Inputs | Text, Images, Audio, Video | Text, Images |
| Weights | Open weights · Apache-2.0 | Open weights · Apache-2.0 |
| Released | Nov 20, 2026 | Nov 20, 2026 |