Roboflow

Muse Spark 1.2 vs Muse Spark Open

Muse Spark 1.2 and Muse Spark Open score 73.8% vs 69.4% averaged across five Vision Evals tasks (#26 and #33 of 73 models), so Muse Spark 1.2 leads overall.

  • Muse Spark 1.2 wins all 5 tasks, with its widest lead on reasoning (75.1% vs 63.8%).
  • Muse Spark Open is cheaper: $0.0059 vs $0.0083 overall average cost per sample.
  • Muse Spark Open is faster: 7.4s vs 11.5s per sample.

Compare Muse Spark 1.2 vs Muse Spark Open live

Run the same image across every model that supports a task and compare their outputs side-by-side.

Detect and compare bounding boxes across models on the same image.

Open Object Detection in the full playground
MetaMuse Spark 1.2
Run to compare this model.
MetaMuse Spark Open
Run to compare this model.

Muse Spark 1.2 vs Muse Spark Open on Vision Evals

Score by taskMuse Spark 1.2Muse Spark Open
Reasoning
75.1
63.8
Identification
89.6
82.3
Object Detection
59.0
57.8
OCR
68.7
67.6
Counting
76.6
75.7

Roboflow Vision Evals. Full breakdown below.

Muse Spark 1.2 vs Muse Spark Open Comparison Table

Evals updated November 19, 2026Pricing updated November 21, 2026

PropertyMuse Spark 1.2Muse Spark Open
OrganizationMetaMeta
Categoryclosedopen
Modalitymultimodalmultimodal
Release DateAug 2026Oct 2026
Context Window1.0M1.0M
ParametersUnknown816B total / 42B active
LicenseProprietaryCustom
Commercial usecommercial use per vendor termsSee license terms
Pricing per 1M tokens
Input $/1M$1.25$1.25
Output $/1M$4.25$4.25
Vision Tasks
Both support 10 more tasks
Model Features
Both support 3 more features
Vision Evalsground-truth scores across 5 vision tasks, pooled at low effort
Overall
73.8%
69.4%
Avg cost / sample$0.0083$0.0059
Avg speed / sample11.51s7.37s
By task
Object Detection (low)
59.0%
±1.0, Mean of 3 runs, range 58.1 to 60.2
$0.0096
57.8%
±0.6, Mean of 3 runs, range 57.0 to 58.2
$0.0073
Object Detection (high)
60.5%
±0.3, Mean of 3 runs, range 60.2 to 60.7
$0.014
60.9%
±0.8, Mean of 3 runs, range 60.3 to 61.9
$0.013
Counting (low)
76.6%
±2.7, Mean of 3 runs, range 74.3 to 79.7
$0.0050
75.7%
±2.7, Mean of 3 runs, range 73.0 to 78.4
$0.0036
Counting (high)
75.2%
±2.0, Mean of 3 runs, range 73.0 to 77.0
$0.0082
75.2%
±4.1, Mean of 3 runs, range 71.6 to 79.7
$0.0073
Identification (low)
89.6%
±1.6, Mean of 3 runs, range 87.5 to 90.6
$0.0038
82.3%
±1.6, Mean of 3 runs, range 81.3 to 84.4
$0.0027
Identification (high)
87.5%
±0.0, Mean of 3 runs, range 87.5 to 87.5
$0.0062
80.2%
±3.1, Mean of 3 runs, range 78.1 to 84.4
$0.0056
OCR (low)
68.7%
$0.0086
67.6%
$0.0061
by category
Single value
60.4%
Transcription
89.3%
Structured JSON
87.0%
Text localization
31.5%
Single value
60.0%
Transcription
89.0%
Structured JSON
85.5%
Text localization
29.5%
OCR (high)
68.6%
$0.014
68.6%
$0.014
by category
Single value
61.7%
Transcription
85.1%
Structured JSON
86.7%
Text localization
30.7%
Single value
58.3%
Transcription
91.4%
Structured JSON
88.2%
Text localization
32.2%
Reasoning (low)
75.1%
±0.3, Mean of 3 runs, range 74.8 to 75.5
$0.0073
63.8%
±1.0, Mean of 3 runs, range 62.9 to 64.9
$0.0049
Reasoning (high)
75.7%
±0.3, Mean of 3 runs, range 75.5 to 76.2
$0.012
65.1%
±3.0, Mean of 3 runs, range 62.3 to 68.2
$0.012

Muse Spark 1.2 vs Muse Spark Open: Overview

Muse Spark 1.2

Muse Spark 1.2 is a proprietary multimodal reasoning model from Meta Superintelligence Labs, released as a coding-focused update to Muse Spark 1.1. It accepts text, images, video, audio, and PDF documents and returns text, with a context window of roughly one million tokens that allows whole repositories, long documents, and extended agent trajectories to be held in a single request. The model thinks before answering, and the amount of reasoning effort it spends is configurable per request. Alongside its visual and document understanding, it supports structured output and parallel function calling, and it is designed to operate either as a planning agent that delegates work or as a subagent executing tasks in parallel.

Training for version 1.2 scaled up compute on coding tasks and widened the diversity of training environments, concentrating on long-horizon work such as whole-repository generation, large end-to-end projects, and automated research. Part of the training data was self-generated, with Muse Spark 1.1 producing coding environments and instruction-following templates and grading candidate solutions against them. The model was co-trained with the Muse Code terminal agent, incorporating rejection-sampled harness trajectories and that toolset. Multimodal use cases documented for the family include visual-to-code generation and detailed image and video captioning.

Muse Spark Open

Muse Spark Open is an open-weight multimodal model released by Meta on October 15, 2026. Its sparse mixture-of-experts architecture has 816 billion total parameters and activates 42 billion per token. It accepts text, images, audio and video, produces text, and supports a context window of 1,048,576 tokens. Meta provides BF16 and NVFP4 checkpoints for self-hosting, alongside access through hosted inference partners.

The model is designed for coding, tool use and multi-step reasoning over long tasks. Its visual inputs support workflows involving screenshots, documents and other image content. Vision Evals measures its text understanding, counting, identification, reasoning and object detection at Low and High effort. The published weights use the custom Muse Model License, with commercial use subject to its terms and the accompanying usage policy.