Muse Spark 1.2 Pricing, Benchmarks & Channels
Muse Spark 1.2 is a 2026-08-05 model from Meta featuring a 1M context window and extended reasoning capabilities. Lowest rate on Meta Official API ($1.25 in / $4.25 out per 1M tokens).
Input Price / 1M
$1.25
Output Price / 1M
$4.25
Cache Read / 1M
$0.15
Context Window
1M
Generation Speed
154 TPS
Access Grade
A · Global
F.A.R.E. Four-Dimension Score Breakdown
Scenario Capability Breakdown (0–100):
💻 Coding
88/100
✍️ Writing
87/100
📚 Long-Doc RAG
94/100
👁️ Vision OCR
88/100
F.A.R.E. CAPABILITY RADAR
Muse Spark 1.2 Multi-Platform Price Comparison (2 Verified Channels)
All prices are in USD per 1M tokens. Domestic developers can prioritize channels marked "Direct in CN" for compliant access and local billing.
| Platform / Gateway | Network Access | Input ($/1M) | Output ($/1M) | Cache Hit ($/1M) | Billing & Notes | Link |
|---|---|---|---|---|---|---|
| Meta Official API Lowest Price | Global Network | $1.25 | $4.25 | $0.15 | Global cloud billing | Visit Official → |
| OpenRouter | Direct in CN | $1.25 | $4.25 | $0.15 | Direct CN access · RMB billing | Visit Official → |
✅ Core Strengths & Ideal Workloads
- Top-tier empirical capability score (88 index)
- Fast generation speed (~154 tokens/sec)
⚠️ Limitations & Practical Caveats
- Meta Muse and Llama families are widely hosted across OpenRouter, Together AI, Fireworks, and Bedrock.
- Extended thinking generates internal reasoning tokens (~1.8x output multiplier)