Muse Spark 1.3 and 1.2 Contributor for JanitorAI: Free Proxy Setup on Composite

At a glance

Use it as a JanitorAI proxy

1. Create an API key on the API Keys page and link Discord on your account page (not needed with credits, a plan or your own provider key).
2. In JanitorAI, open Settings → API, choose OpenAI as the API type, and paste the full URL https://composite.lucidity.sh/v1/chat/completions into the OpenAI API URL field.
3. Paste the key, then type meta/muse-spark-1.3-contributor:free as the model name.
4. Send a short message to check the connection before you load a long chat.

Muse Spark 1.3 Contributor is the cost-efficient contributor tier of Meta's multimodal reasoning model, aimed at experimentation, learning, and early agent, multi-agent and coding work. It reads text, images, video and files, and has a window of just over a million tokens. Muse Spark 1.2 Contributor, the earlier version, is also on the free list as meta/muse-spark-1.2-contributor:free.

It is moderated

The provider behind both Muse Spark models moderates requests. Prompts or replies its filter rejects come back as errors instead of text, so mature roleplay will not run on them whatever preset you use. Amazon Nova Micro is the only other moderated model on the free list. For those chats, pick a model such as MiMo-V2.6-Flash or DeepSeek V4 Flash 0731.

It always reasons

Thinking cannot be turned off, but you can keep it short. Add an effort to the model name: meta/muse-spark-1.3-contributor:free[minimal] for the quickest replies, or [high] for harder problems. Version 1.3 also offers [max]; 1.2 stops at [xhigh].

A short list of settings

Both versions accept temperature, top P, top K, repetition penalty, max tokens, tools and structured output. Min P, frequency and presence penalties and stop sequences are not supported, so presets built around them will not have their usual effect.

1.3 or 1.2?

The two cost the same share of the daily allowance and take the same settings. Use 1.3, the newer version, unless you are comparing against results you already had from 1.2.

Other chat apps and the API

SillyTavern, ChubAI, Agnai and RisuAI take the same key and model name with the shorter base URL https://composite.lucidity.sh/v1.

Step-by-step setup for each app is in the JanitorAI and SillyTavern guides, and the docs hub covers the others.

From code, it is a normal OpenAI-compatible request:

curl https://composite.lucidity.sh/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "meta/muse-spark-1.3-contributor:free", "messages": [{"role": "user", "content": "Hello"}]}'

Free access (as of November 1, 2026)

On November 1, 2026, this model is on Composite's free list as meta/muse-spark-1.3-contributor:free. The plain meta/muse-spark-1.3-contributor stays paid. Every account shares 80 requests a day across all of the free models, separate from Synth's count, and the count resets at midnight UTC. Pricier free models count as several requests per call; the Models page shows how many. Credits and plans do not raise it.

The free list changes. Check the free models on the Models page for the current set, and the features guide for the rules. A :free id that is no longer on the list returns a model-not-found error and charges nothing. When the 80 requests are used up, requests stop with an error until the reset; Composite never charges your credits to continue. You can switch to meta/muse-spark-1.3-contributor on credits or a plan, or to Synth.

More

• Model page with live pricing and rankings
• Model modifiers, including reasoning effort suffixes
• Best LLMs for Roleplay in 2026
• All model guides