A 55B-parameter mixture-of-experts model from NVIDIA with a 262K token context window, using a hybrid Transformer-Mamba architecture for reasoning and task orchestration.
- Strengths
- Efficiently routes computation across expert modules, enabling strong performance on complex reasoning tasks without requiring full model activation across all parameters.
- Best for
- Multi-step reasoning, task planning, and workflows that benefit from sparse activation and a long context window for handling complex documents or conversations.
- Limitations
- MoE architecture requires hardware support for efficient inference; performance on tasks that don't benefit from expert specialization or routing may not justify the added complexity.
Input / 1M
$0.5
Output / 1M
$2.20
Cached input / 1M
$0.1
Context window
262K
Price history
Input (solid)Output (dashed)
Price change
- 30d
- in decreased 0.0% out decreased 0.0%
- 90d
- in decreased 0.0% out decreased 12.0%
- 1y
- in decreased 0.0% out decreased 12.0%
- Since launch
- in decreased 0.0% out decreased 12.0%
Snapshots
| Effective | Input | Output | Cached in | Note | Source |
|---|---|---|---|---|---|
| 28 Jul 2026 | $0.5 | $2.20 | $0.1 | Imported from OpenRouter | openrouter.ai |
| 27 Jul 2026 | $0.5 | $2.20 | $0.1 | Imported from OpenRouter | openrouter.ai |
| 25 Jul 2026 | $0.6 | $3.60 | $0.2 | Imported from OpenRouter | openrouter.ai |
| 23 Jul 2026 | $0.5 | $2.20 | $0.1 | Imported from OpenRouter | openrouter.ai |
| 15 Jul 2026 | $0.6 | $3.60 | $0.2 | Imported from OpenRouter | openrouter.ai |
| 17 Jun 2026 | $0.5 | $2.20 | $0.1 | Imported from OpenRouter | openrouter.ai |
| 11 Jun 2026 | $0.5 | $2.50 | $0.15 | Imported from OpenRouter | openrouter.ai |