← All Models
Muse Spark 1.3
Meta's general agent model for reasoning, coding, research, and multimodal understanding. Text, image, video, and PDF input; text output. One 1,048,576-token window is shared by input, reasoning, and output. Always-on reasoning from minimal to max. Function calling and native web search.
Web Search
Specifications
| Provider | Meta |
| Category | Text |
| Context window | 1.0M tokens |
Pricing (Starter tier)
162.50 credits / 1M input tokens
19.50 credits / 1M cached input tokens
552.50 credits / 1M output tokens
325.00 credits / 1K web searches
Higher tiers get volume discounts. See tiers
Quick Start
curl -N https://query.genx.sh/v1/chat/completions \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"muse-spark-1.3","messages":[{"role":"user","content":"Compare three approaches to caching an API."}],"reasoning_effort":"medium","stream":true}'
Parameters
| Name | Type | Required | Default | Description |
|---|---|---|---|---|
prompt |
string | required | — | Input text prompt |
system_prompt |
string | optional | — | System instructions |
max_tokens |
number | optional | — | 16 to 1048576 output tokens including reasoning; input and output share the context window |
reasoning_effort |
string | optional | — | minimal, low, medium, high, xhigh, or max; reasoning cannot be turned off |
tools |
array | optional | — | Built-in web_search |
Availability
Muse Spark 1.3 is available in the GenX Pro App and via the GenX Router API.