Muse Spark 1.1

Name: Muse Spark 1.1
Author: Meta

Multimodal

Meta

Muse Spark 1.1 is Meta's first model in the Muse Spark family, designed for agentic and multimodal tasks with a 1M token context window. It leads the MCP Atlas benchmark and ranks highly on OSWorld-Verified and Terminal-Bench 2.1.

Key Specifications

Parameters

Context

1.0M

Release Date

July 9, 2026

Average Score

70.7%

API Documentation Research Paper Results Blog

Timeline

Key dates in the model's history

Announcement

July 9, 2026

Last Update

August 27, 2026

Today

September 8, 2026

Technical Specifications

Parameters

Training Tokens

Knowledge Cutoff

Family

Capabilities

MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)

$1.25

Output (per 1M tokens)

$4.25

Max Input Tokens

1.0M

Max Output Tokens

131.1K

Supported Features

Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Other Tests

Specialized benchmarks

CharXiv-R

1,000 CharXiv validation chart-reasoning questions; gpt-oss-120b high as judge. • Self-reported

88.0%

MCP Atlas

Results sourced from Scale AI. • Self-reported

88.0%

OSWorld-Verified

361-task split excluding Google Drive tasks; GUI-only agent; one attempt per task. • Self-reported

81.0%

Terminal-Bench 2.1

Bash-tool-only harness; pass@1 averaged over 5 attempts on 89 tasks. • Self-reported

80.0%

BabyVision

388-question visual-understanding benchmark; gpt-oss-120b high as judge for free-form answers. • Self-reported

76.3%

DeepSWE 1.1

DeepSWE v1.1 leaderboard, mini-swe-agent harness, xhigh effort; Pass@1 53% ± 3%. • Verified

53.0%

Finance Agent v2

Results sourced from Vals AI. • Self-reported

57.2%

Humanity's Last Exam

With tools. • Self-reported

62.1%

Job Bench

OpenCode harness with Grok 4.3 as judge; results provided by the benchmark authors. • Self-reported

54.7%

SWE-Bench Pro

Scale AI leaderboard, mini-swe-agent harness. • Self-reported

61.5%

Toolathlon

Toolathlon-Verified, official harness. • Self-reported

75.6%

License & Metadata

License

proprietary

Announcement Date

July 9, 2026

Last Updated

August 27, 2026

Compare Muse Spark 1.1

All comparisons

vs Muse Spark 1.2 vs Muse Spark 1.3 vs Muse Glimmer-30B vs Qwen3.8 Max vs Qwen3.8 Flash vs Gemini 3.6 Flash vs Kimi K3 vs Qwen3.8 Flash Next

Similar Models

All Models

Muse Spark 1.2

Meta

Released:Aug 2026

Price:$0.10/1M tokens

Muse Spark 1.3

Meta

Released:Sep 2026

Price:$0.10/1M tokens

Muse Spark

Meta

Best score:0.9 (TAU)

Released:Apr 2026

Llama 3.2 90B Instruct

Meta

MM90.0B

Best score:0.9 (MMLU)

Released:Sep 2024

Price:$1.20/1M tokens

Llama 4 Maverick

Meta

MM400.0B

Best score:0.9 (MMLU)

Released:Apr 2025

Price:$0.27/1M tokens

Llama 3.2 11B Instruct

Meta

MM10.6B

Best score:0.7 (MMLU)

Released:Sep 2024

Price:$0.18/1M tokens

Llama 4 Scout

Meta

MM109.0B

Best score:0.8 (MMLU)

Released:Apr 2025

Price:$0.18/1M tokens

Muse Glimmer-30B

Meta

MM29.6B

Best score:0.8 (GPQA)

Released:Aug 2026

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.