Head to head
Meta: Muse Spark 1.2 vs Anthropic: Claude Sonnet 5
Live prices, context windows, capabilities, and measured quality — updated continuously.
Side by side
| Metric | Meta: Muse Spark 1.2 Meta | Anthropic: Claude Sonnet 5 Anthropic |
|---|---|---|
| Input $/M tokens | $1.25 | $2.00 |
| Output $/M tokens | $4.25 | $10.00 |
| Context window | 1.0M | 1.0M |
| Max output | — | 128K |
| WM Score | 87 | 87 |
| Benchmark quality | 88 | 85.5 |
| Hosts serving it | 1 | 5 |
| Throughput (tok/s) | 129.5 | 70 |
| Latency (ms) | 6768 | 1325 |
| Vision | ||
| Tools | ||
| Reasoning | ||
| JSON mode |
Which should you pick?
Meta: Muse Spark 1.2 costs less per million input tokens — about 1.6× cheaper than Anthropic: Claude Sonnet 5. Meta: Muse Spark 1.2 scores higher on measured quality. Meta: Muse Spark 1.2 takes a larger context window, which matters for long documents or big codebases. Prices here are live list prices; the same model can cost less on a different host, so check the model page for the cheapest offering.