M
Muse Spark OpenOPEN
Muse Spark Open is a coding-optimized open-weight model.
Prices update automatically — checked daily against provider list prices.
- $2 per 1M tokens, assuming 3 input for 1 output the convention used by Artificial Analysis
- cheaper than 5% of the 197 open-weight models tracked here
- 15 of 16 Meta models tracked here, from cheapest
Benchmarks & performance
Independent benchmark scores for Muse Spark Open, measured by Artificial Analysis. Higher is better. Measurement mode: Xhigh.
For the same budget, Muse Spark Open delivers 2.7 times more capability. Open weights also mean you can self-host it and pay nothing per token.
Models at this level cost $1.65 per 1M tokens (median of 23) — this one costs $2.
Cheaper and better on this index: MiMo-V2.6-Pro · GLM 5.3 · GLM 5.3 (batch) and 12 more
About this model
Muse Spark Open is an open-weight AI model by Meta. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Frequently asked questions
What is Muse Spark Open?
Muse Spark Open is an AI language model from Meta. It is open-weight: you can download it and run it on your own hardware, for free. It scores 38.9 on the Artificial Analysis intelligence index.
Is Muse Spark Open free?
The weights are free and open — you can self-host Muse Spark Open and pay nothing per token. If you prefer a hosted API, list prices are $1.25 per million input tokens and $4.25 per million output tokens.
What is Muse Spark Open good at?
Independent benchmarks from Artificial Analysis give it GPQA 89%, Humanity's Last Exam 43.9%, Long Context Reasoning 78.3%, SciCode 58.4%, Terminal-Bench 76.4%.
How fast is Muse Spark Open?
It generates about 104.8 tokens per second, with a median 0.48s delay before the first token. Measured independently by Artificial Analysis.
Can I self-host Muse Spark Open?
Yes. Muse Spark Open has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.