Skip to main content
Core.Today
|
Billing note Reasoning model: a 3-word reply used ~700 completion tokens (687 reasoning) in testing, and max_tokens 400 ended with empty content. Use max_tokens >= 2048; reasoning tokens are billed at the output rate.
MetaMediumUltra

Muse Spark 1.3

Meta's Muse Spark 1.3 over the direct Meta API - a multimodal reasoning model for agentic work that accepts text, images, video, audio and PDF documents and returns text, with a 1M-token context window and cached-input pricing. Answers are generated after reasoning tokens, so keep max_tokens at 2,048 or more.

2,323/7,897credits
input / output ยท per 1M tokens
1M-token context window
Multimodal input - text, image, video, audio, PDF
Reasoning model built for agentic tasks
Cached input pricing (~88% discount)
OpenAI-compatible API (chat completions) via /llm/meta/v1
Billed exactly at the listed rates (direct route, not an aggregator)

Run it right now

Test this model instantly in the Console Playground โ€” no code required

Sign in to try

Use with AI Assistant

Copy usage instructions for Claude, ChatGPT, or other AI

Model Specifications

Context Window
1.0M
tokens
Max Output
33K
tokens
Training Cutoff
Not published
Compatible SDK
OpenAI

Capabilities

Vision
Function Calling
Streaming
JSON Mode
System Prompt

Token Pricing (per 1M tokens)

Token TypeCreditsUSD Equivalent
Input Tokens2,323$1.55
Output Tokens7,897$5.26
Cached Tokens278.7$0.19

* 1,500 credits โ‰ˆ $1 (actual charges may vary based on usage)

Quick Start

curl -X POST "https://api.core.today/llm/meta/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer cdt_your_api_key" \
  -d '{
  "model": "muse-spark-1.3",
  "messages": [
    {
      "role": "user",
      "content": "Summarize the attached report in 5 bullets."
    }
  ],
  "max_tokens": 4096
}'

Parameters

ParameterTypeRequiredDefaultDescription
messagesarrayYes-Chat messages in OpenAI format (system/user/assistant roles).
max_tokensintegerNo2048Maximum completion tokens including reasoning. Keep >= 2048 or the answer may be cut off before any content is produced.
temperaturenumberNo0.7Sampling temperature (0-2).
streambooleanNofalseStream the response as server-sent events.

Examples

Chat completion

OpenAI-compatible chat completion over the direct Meta route

curl -X POST "https://api.core.today/llm/meta/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer cdt_your_api_key" \
  -d '{
  "model": "muse-spark-1.3",
  "messages": [
    {
      "role": "user",
      "content": "Summarize the attached report in 5 bullets."
    }
  ],
  "max_tokens": 4096
}'

Tips & Best Practices

1Works with the OpenAI SDK - set base_url to https://api.core.today/llm/meta/v1
2Keep max_tokens >= 2048: reasoning tokens come first and are billed at the output rate
3For discounted pricing where prompts may be used for Meta training, see muse-spark-1.3-contributor
4Repeated system prompts and documents benefit from the cached-input rate

Use Cases

Agentic workflows over mixed media (documents, screenshots, recordings)
Video and audio understanding with text answers
Long-context reasoning across large PDF collections