meta
Meta: Muse Spark 1.2
Meta's Muse Spark 1.2 is an AI model capable of processing text, images, videos, files, and audio, with a vast context length of 1,048,576 tokens. It supports tools and reasoning, making it versatile for complex tasks; however, structured output is not a feature. This model should be considered by those requiring a robust suite of input modalities and reasoning capabilities, despite its premium pricing of $1.25 per million input tokens and $4.25 per million output tokens. Its benchmark blended score of 92.4 across three independent benchmarks is strong, but the limited coverage might indicate unproven performance in certain areas.
Quality Score
100/100
price + capability + benchmarks
Input Price
$1.25
per 1M tokens
Output Price
$4.25
per 1M tokens
Context Window
1,048,576
tokens
Benchmark results
Independent, published benchmarks. Blended score 92.4 across 3 benchmarks, last refreshed 2026-08-13. How scoring works →
| Benchmark | Measures | Score |
|---|---|---|
| AI Index | broad capability composite | 93.7 |
| AI Index Coding | software engineering tasks | 100.0 |
| AI Index Agentic | multi-step tool-using tasks | 81.4 |
- Model ID
- meta/muse-spark-1.2
- Vendor
- meta
- Released
- August 2026
- Tokenizer
- Other
- Input Modalities
- text, image, video, file, audio
- Output Modalities
- text
- Max Output
- default
- Tool Calling
- ✓ supported
- Structured Output
- ✓ supported
- Reasoning Mode
- ✓ supported
- Vision
- ✓ accepts images
- Audio
- ✓ accepts audio
- Moderated
- yes
What it costs in practice
Computed from the current $1.25/M input and $4.25/M output rates. Run your own numbers →
| Job | Tokens | Cost |
|---|---|---|
| Summarize a 50-page report | 30k in / 1.5k out | $0.04 |
| Classify 1,000 customer emails | 500k in / 50k out | $0.84 |
| A month of a busy support chatbot | 5M in / 2M out | $14.75 |
Strong choice for
Category rankings
Where Meta: Muse Spark 1.2 places across the 56 categories it ranks in. How we rank →
| # | Category | Score |
|---|---|---|
| #1 | TTS ReplacementVoice · of 25 ranked | 115 |
| #3 | Audio SummarizationVoice · of 25 ranked | 161 |
| #5 | Video SummarizationVideo · of 25 ranked | 161 |
| #9 | Resume WritingBusiness · of 25 ranked | 118 |
| #10 | Sales / Cold EmailBusiness · of 25 ranked | 124 |
| #10 | Job Application DraftingBusiness · of 25 ranked | 124 |
| #10 | Journaling HelperPersonal · of 25 ranked | 124 |
| #10 | Recipe GenerationPersonal · of 25 ranked | 124 |
| #12 | Regex WritingCode · of 25 ranked | 137 |
| #12 | Language TranslationWriting · of 25 ranked | 137 |
| #12 | Tutoring for KidsPersonal · of 25 ranked | 137 |
| #13 | SOP / Process DocsBusiness · of 25 ranked | 138 |
| #14 | Diagram ExtractionVision · of 25 ranked | 150 |
| #14 | Screenshot DebuggingVision · of 25 ranked | 144 |
| #14 | Chart & Graph ReadingVision · of 25 ranked | 153 |
| #14 | Function / Tool CallingAgents · of 25 ranked | 155 |
| #16 | Frontend Component DesignCode · of 25 ranked | 147 |
| #16 | OCR / Document ParsingData · of 25 ranked | 147 |
| #16 | Table Extraction from PDFsData · of 25 ranked | 147 |
| #16 | Marketing CopyWriting · of 25 ranked | 131 |
| #16 | Standardized Test PrepEducation · of 25 ranked | 131 |
| #17 | Blog Post WritingWriting · of 25 ranked | 141 |
| #17 | Browser AutomationAgents · of 25 ranked | 173 |
| #17 | Trip PlanningPersonal · of 25 ranked | 141 |
| #18 | SQL GenerationCode · of 25 ranked | 173 |
| #19 | Code DocumentationCode · of 25 ranked | 143 |
| #19 | Fitness CoachingPersonal · of 25 ranked | 136 |
| #20 | CSV / Spreadsheet CleanupData · of 25 ranked | 154 |
| #20 | TranscriptionVoice · of 25 ranked | 115 |
| #20 | Character RoleplayPersonal · of 25 ranked | 151 |
| #21 | Unit Test GenerationCode · of 25 ranked | 155 |
| #21 | Data AnalysisData · of 25 ranked | 173 |
| #21 | Long-Document SummarizationWriting · of 25 ranked | 161 |
| #21 | Financial AnalysisProfessional · of 25 ranked | 161 |
| #21 | Math TutoringEducation · of 25 ranked | 149 |
| #21 | Physics TutoringEducation · of 25 ranked | 149 |
| #22 | Bug FixingCode · of 25 ranked | 182 |
| #22 | Scientific CodingResearch · of 25 ranked | 182 |
| #23 | Agent WorkflowsAgents · of 25 ranked | 194 |
| #23 | Coding AgentsAgents · of 25 ranked | 194 |
| #23 | Experiment DesignResearch · of 25 ranked | 159 |
| #24 | Code RefactoringCode · of 25 ranked | 165 |
| #24 | Creative WritingWriting · of 25 ranked | 159 |
| #24 | Screenwriting & DialogWriting · of 25 ranked | 159 |
| #24 | Academic WritingWriting · of 25 ranked | 159 |
| #24 | Legal DraftingProfessional · of 25 ranked | 159 |
| #24 | Legal ResearchProfessional · of 25 ranked | 169 |
| #24 | Scientific ResearchProfessional · of 25 ranked | 169 |
| #24 | History TutoringEducation · of 25 ranked | 146 |
| #24 | OKRs & Strategic PlanningBusiness · of 25 ranked | 159 |
| #24 | RAG PipelinesAgents · of 25 ranked | 165 |
| #24 | Long-Context Q&AAgents · of 25 ranked | 155 |
| #24 | Fiction CollaboratorPersonal · of 25 ranked | 169 |
| #24 | Math ProofsResearch · of 25 ranked | 159 |
| #24 | Literature ReviewResearch · of 25 ranked | 169 |
| #25 | CI/CD PipelinesCode · of 25 ranked | 145 |
Similar models
Quick answers
- How much does Meta: Muse Spark 1.2 cost?
- $1.25 per million input tokens and $4.25 per million output tokens.
- What is Meta: Muse Spark 1.2's context window?
- 1,048,576 tokens, roughly 1,572 pages of text in a single request.
- Does Meta: Muse Spark 1.2 support tool calling?
- It supports tool calling, structured output, a reasoning mode.
- Can Meta: Muse Spark 1.2 process images?
- Yes, it accepts image input.