Meta の Muse Spark 1.1 は、テキストと画像のネイティブ・マルチモーダル理解および図表・ドキュメント解析向けに設計されたAIモデルです。1M のコンテキストウィンドウ、毎秒 105 トークンの生成速度、知能指数 90/100（1Mトークンあたり入力$3・出力$15）を実現しています。
Technical Specifications
- Provider
- Meta
- Category
- flagship
- Release Date
- 2026-04
- Batch API Discount
- 50% off async batch
- Consumer Subscription
- $20 / month
- Reasoning Effort Control
- Configurable (Low / Med / High)
- Structured Outputs
- Yes
- Output Modalities
- text
- Official Documentation
- Visit Official Docs ↗
Data Privacy, Licensing & Deployment Governance
Official Provider Release Intelligence, API Quotas & Agentic Specs
- Official API Model ID (SDK String)
- meta-llama/Muse-Spark-1.1
- Supported SDK & API Protocol
- Meta REST API (OpenAI-Compatible)
- Lifecycle & Deprecation Status
- Active GA (Guaranteed support through 2028+)
- Production Rate Limits (RPM & TPM)
- 5,000 – 15,000 RPM (Production Tier) · 4,000,000 – 10,000,000 TPM
- Long-Context Pricing Tier (>128K / >200K)
- Flat rate ($3/1M in) across full 1000K window
- Prompt Cache Write & TTL Policy
- Cached input at $0.3/1M tokens
- Computer Use & Browser GUI Agent
- Supported (Screen coordinate grounding, browser & desktop GUI action loops)
- MCP & Parallel Tool Calling
- Native MCP Client/Server ready · Parallel function calling · 100% Strict JSON Schema
- Web Search Grounding & Code Sandbox
- Supported via external Tavily/Exa Search & E2B/Docker Sandbox tools
- Fine-Tuning & Distillation Support
- API Prompt Engineering & RAG Distillation
- Multimodal Tokenization Billing
- Text BPE (~4 chars/token) · Vision (~258–765 tokens per image tile) · Audio (~25–32 tokens/sec) · Video (~258–300 tokens/sec)
- Weights Precision & Quantization
- Cloud-Hosted Proprietary (FP8 / BF16 Tensor Core Cluster)
- Minimum GPU VRAM Requirement
- Zero Local VRAM Required (100% Serverless Cloud API)
- Supported Inference Engines
- Managed Meta Cloud API
- System Card & Safety Framework
- Meta Production Safety & Alignment Verified
- Enterprise Compliance (SOC2 / HIPAA / ZDR)
- SOC2 Type II · GDPR · Zero Data Retention (ZDR) & HIPAA BAA Available on Enterprise Tier
Muse Spark 1.1 Cost, Context, Hardware & API Integration
For single-user chat, a $20/mo subscription caps your bill below this API volume.
Break-even point: ~31 requests/day equals $20/month.
Same Monthly Workload Cost Across Popular Reference Models
Live Prompt Paste Tokenizer & Cost Counter
Paste real system instructions, JSON schemas, or code to measure BPE token counts across tokenizers.
Hybrid Prompt Routing Cascade Simulator
Route routine queries to a fast lightweight model ($0.15/1M) and escalate only complex reasoning turns to Muse Spark 1.1:
Benchmark Scores & 6-Axis Capability Radar
- SWE-bench Verified
- 76.5%
- GPQA Diamond
- 83.5%
- MMLU-Pro
- 88%
- HumanEval
- 95%
- Chatbot Arena Elo
- 1495pts
Strengths
- 1M のコンテキストウィンドウと最大 64K トークンの出力容量（SWE-bench Verified 76.5% · GPQA Diamond 83.5%）
- 毎秒 105 トークンの高速スループットと、初回トークン到達時間（TTFT）わずか 600ms の低遅延
- 優れたAPI経済性（1Mトークンあたり入力$3・出力$15）に加え、プロンプトキャッシュや構造化JSON出力・関数呼び出しに完全対応
Limitations & Trade-offs
- クラウド専用のプロプライエタリモデルのため、モデル重みをダウンロードしてオフライン実行することはできません
- 大量のトークンを処理する本番環境では、プロンプトキャッシュや Batch API の併用によるコスト最適化が推奨されます
Editorial Assessment
知能指数 90/100、1M コンテキスト、毎秒 105 トークンの速度を兼ね備えた Muse Spark 1.1 は、Meta のラインナップにおいて非常に実用性の高い主力モデルです。
Best for: テキストと画像のネイティブ・マルチモーダル理解および図表・ドキュメント解析、エンタープライズRAG基盤、スケーラブルな本番API統合に最適。
Compare Muse Spark 1.1
Frequently Asked Questions about Muse Spark 1.1
How much does the Muse Spark 1.1 API cost per 1M tokens?+
Muse Spark 1.1 by Meta is priced at $3 per 1M input tokens and $15 per 1M output tokens. With prompt caching enabled, cached input tokens cost $0.3 per 1M tokens, and asynchronous Batch API requests receive a 50% discount.
Should I use a $20/month subscription or the Muse Spark 1.1 pay-as-you-go API?+
A flat $20/month consumer subscription is predictable for daily interactive chat and manual coding. However, if you are building software, running automated workflows, or sending low-to-moderate volume with prompt caching enabled, the pay-as-you-go API is typically far more cost-effective and offers higher rate limits without manual UI caps.
How many words and pages fit in Muse Spark 1.1's context window?+
Muse Spark 1.1 features a 1M tokens context window, which equals approximately 750,000 English words or ~2,000 standard single-spaced pages in a single prompt. Its maximum output generation per turn is 64,000 tokens.
Can I run Muse Spark 1.1 locally on my own GPU (Open Weights)?+
No, Muse Spark 1.1 is a proprietary closed-weights model hosted by Meta. It cannot be downloaded for local GPU execution and must be accessed via Meta's official API or cloud platform.
Does Meta train on data sent to the Muse Spark 1.1 API, and is commercial use allowed?+
API payloads are excluded from model training by default. License tier: Proprietary Commercial API (Meta Terms). Full commercial ownership of generated outputs. Supported deployment channels include: Meta Cloud API & Enterprise Endpoints.
How does Muse Spark 1.1 perform on coding and reasoning benchmarks?+
Muse Spark 1.1 achieves a composite Intelligence Index of 90/100, scoring 76.5% on SWE-bench Verified, 83.5% on GPQA Diamond, 88% on MMLU-Pro, and 1495 pts on Chatbot Arena Elo.
What is Muse Spark 1.1 best used for?+
Muse Spark 1.1 is best suited for: テキストと画像のネイティブ・マルチモーダル理解および図表・ドキュメント解析、エンタープライズRAG基盤、スケーラブルな本番API統合に最適。. Assessment: 知能指数 90/100、1M コンテキスト、毎秒 105 トークンの速度を兼ね備えた Muse Spark 1.1 は、Meta のラインナップにおいて非常に実用性の高い主力モデルです。
Meta Family Lineage, Routing Tiers & Upgrade Path
Step down to a lighter sibling for high-volume routing or step up to Meta's flagship reasoning tier:
Llama 4 Behemoth
Llama 5
Llama 4 Maverick
Llama 4 Scout
Llama 3.3 70B
Llama 3.2 11B Vision
AI Glossary & Architecture Concepts
View full glossary →Key machine learning concepts and benchmarks relevant to Muse Spark 1.1:
The basic unit of text that a language model processes, typically representing about 3 to 4 English characters or roughly 0.75 words. For example, the sentence "Artificial intelligence is fast" breaks down into about 5 tokens.
A standardized evaluation suite used to objectively measure and rank an AI model's capabilities across specific tasks — such as SWE-bench Verified for software engineering, AIME for competitive math, or MMLU for general knowledge.
The maximum number of tokens a model can consider at once — both the input it reads and the output it writes. A 1,000,000-token context window can hold roughly 750,000 English words in a single request.
An AI model trained via reinforcement learning to spend dynamic compute thinking through complex logic, math, and architectural problems before outputting a solution. Examples include OpenAI o1, o3, and DeepSeek R1.
A step-by-step reasoning technique where an AI model generates intermediate problem-solving thoughts before arriving at a final answer. Native reasoning models like OpenAI o3 and DeepSeek R1 use CoT to solve complex math and coding challenges.
The ability of a language model to recognize when an external capability is needed and output structured JSON arguments to invoke external APIs, databases, or terminal commands. The runtime executes the tool and passes results back to the model.
Benchmark scores, token prices, and rate limits evolve rapidly as AI labs release model updates. Always verify current specifications and pricing on the official provider documentation.