WizardLM-2 8x22B
VerifiedMicrosoft's open-weight MoE model for complex instruction tasks.
About WizardLM-2 8x22B
WizardLM-2 8x22B builds on the Mixtral 8x22B architecture and applies Microsoft's WizardLM fine-tuning approach. This results in an open-weight model that users can download and run locally or on their own infrastructure. The 65536-token context window enables processing of lengthy documents and extended conversations.
Its strengths include handling detailed instructions and producing coherent multi-turn responses across varied text domains. Because the weights are publicly available, researchers and developers can fine-tune the model further for specialized needs without relying on proprietary APIs.
Typical usage covers chat interfaces, code assistance, summarization of long texts, and exploratory research projects. The text-only modality keeps it focused on language-based workflows rather than multimodal inputs.
Capabilities
How WizardLM-2 8x22B compares
WizardLM-2 8x22B (striped bar) vs other language models on intelligence, speed and price.
Price
USD per 1M output tokens · Lower is better · WizardLM-2 8x22B ranks #56 of 141
Sources: Artificial Analysis (intelligence, speed) · OpenRouter (price).
Best for
Long-Context Document Summarization
The model processes up to 65536 tokens to summarize and analyze lengthy reports or research papers while maintaining coherence across sections.
Complex Codebase Refactoring
It handles multi-step code generation and problem solving by following detailed instructions to refactor large software projects.
Creative Narrative Development
Users can generate extended stories or scripts that require consistent character tracking and plot progression over many tokens.
Strengths & limitations
Strengths
- +Strong instruction adherence
- +Effective extended context handling
- +Versatile reasoning across domains
Limitations
- –Text-only modality
- –High inference compute demands
- –Risk of factual hallucinations
Cost calculator
Estimate what WizardLM-2 8x22B would cost for your usage.
Based on WizardLM-2 8x22B's $0.62/1M input · $0.62/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "microsoft/wizardlm-2-8x22b",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: microsoft/wizardlm-2-8x22b
Editor's verdict
WizardLM-2 8x22B is Microsoft's open-weight language models with a 66K-token context window.
At $0.62 per 1M output tokens, it is very cost-efficient for its class.
As an open-weight model you can self-host it or call it through a hosted API.
Best suited to strong instruction adherence and effective extended context handling.
Frequently asked questions
The model supports a context length of 65536 tokens.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Similar models
Other language models worth comparing.