Back to models
Meta
Muse Spark Open
MoEotherReleased Oct 15, 2026
Meta's open-weight multimodal mixture-of-experts model with 816B total and 42B active parameters, optimized for agentic coding and tool use. Supports a 1,048,576-token context window and can run on self-hosted infrastructure or hosted inference providers. Released under the Muse Model License with limited commercial terms, not Apache 2.0.
ChatText completionFunction callingVisionAudioCode generationReasoningtext · image · audio · video · code
Provider (1)
Sorted by total cost (input + output per 1M tokens). Select a row to view provider details.
| Provider | Pricing (per 1M) | Rate limits | Regions | Health | Latency |
|---|---|---|---|---|---|
 | In: $1.25Out: $4.25 | N/A RPM / N/A TPM | us-east-1 | Healthy | 0ms |
Quick start
Use this model via OpenRouter with an OpenAI-compatible SDK.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const response = await client.chat.completions.create({
model: "meta/muse-spark-open",
messages: [
{ role: "user", content: "Hello!" }
],
});
console.log(response.choices[0].message.content);Using OpenRouter API · OpenAI-compatible SDK