Cosmos Reason2 8B
by Nvidia · listed Dec 2025RTVW
Vision language model for physical-world understanding with structured reasoning on video and images
Current rates USD per 1M tokens
Inputfree
Outputfree
Cache read—
Cache write—
Against the market
Input price vs. maker flagships ($/1M tokens)Output price vs. maker flagships ($/1M tokens)On the record
- Model id
nvidia/cosmos-reason2-8b- Context window
- 131K tokens
- Max output
- 16K tokens
- Input modalities
- text, image, video
- Output modalities
- text
- Reasoning
- yes
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2025-12-01
- Last updated
- 2025-12-01
← All Nvidia listings