Muse Spark Open - Kyma API

Overview

Meta’s open-weight Muse Spark (816B MoE, 42B active), trained for agentic coding: higher first-attempt accuracy, reliable tool calling and deep codebase understanding. 1M context, image input, reasoning.

Specs

Field	Value
Model ID	`muse-spark-open`
Best for	Agentic coding across large repositories on open weights
Context window	1.048576M
Max output tokens	944K
Input modalities	Text, Image
Output modalities	Text
Tool calling	Yes
Structured outputs	Yes
Prompt caching	No
Vision	Yes
Reasoning	Yes
Speed	Medium
Cost band	Premium
Release stage	Stable

Pricing

	Per 1M tokens
Input	$1.688
Output	$5.738

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="kyma-...")

response = client.chat.completions.create(
    model="muse-spark-open",
    messages=[{"role": "user", "content": "..."}],
    tools=[...],  # function calling supported
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=chat&tools=true&quality_tier=frontier-open"

​Overview

​Specs

​Pricing

​Example

​Agent query example

Overview

Specs

Pricing

Example

Agent query example