WizardLM-2 8x22B uses a mixture-of-experts architecture to deliver strong performance while activating only a fraction of parameters per token. It remains competitive for coding and complex instruction-following tasks.
Pricing Breakdown
| Volume | Input Cost | Output Cost | Combined (50/50) |
|---|---|---|---|
| 1,000 tokens | $0.0006 | $0.0006 | $0.0006 |
| 10,000 tokens | $0.0063 | $0.0063 | $0.0063 |
| 100,000 tokens | $0.0630 | $0.0630 | $0.0630 |
| 1,000,000 tokens | $0.6300 | $0.6300 | $0.6300 |
Strengths
- ✓Mixture-of-experts architecture for efficiency
- ✓Open weights
- ✓Strong at complex instruction following
- ✓Good coding performance
Weaknesses
- ✗Large resource requirement to self-host
- ✗Superseded by newer models in most benchmarks
Best For
Frequently Asked Questions
How much does WizardLM-2 8x22B cost?
WizardLM-2 8x22B costs $0.63/1M per 1M input tokens and $0.63/1M per 1M output tokens.
What is WizardLM-2 8x22B's context window?
WizardLM-2 8x22B has a context window of 66K tokens, which means it can process up to 65,536 tokens in a single request.
Is WizardLM-2 8x22B open source?
Yes, WizardLM-2 8x22B is open source. The model weights are publicly available and can be self-hosted.
What is WizardLM-2 8x22B best used for?
WizardLM-2 8x22B is best suited for: coding, creative-writing, research. Mixture-of-experts architecture for efficiency.
What is WizardLM-2 8x22B's ELO score?
WizardLM-2 8x22B has an ELO score of 1190 on the LMSYS Chatbot Arena leaderboard, placing it in the mid-tier range.