🇺🇸 United States
| Model | Coding index | tok/s | $/M in | $/M out | $ / session | Cache |
|---|---|---|---|---|---|---|
| Muse Glimmer | 49.0 | 90.2 | $0.35 | $1.50 | — | — |
| Muse Spark 1.1 | 71.3 | 0.0 | $1.25 | $4.25 | — | — |
| Muse Spark 1.2 | 72.2 | 0.0 | $1.25 | $4.25 | — | — |
| Llama 4 Maverick | 16.3 | 112.9 | $0.20 | $0.80 | — | — |
| Llama 4 Scout | 8.2 | 131.1 | $0.10 | $0.30 | — | — |
2 endpoints, cheapest first. This is the view the model-by-model tables cannot give you: what one provider actually offers, and at what precision.
| Model | Quant | $/M in | $/M out | Context | Uptime 30m |
|---|---|---|---|---|---|
| Muse Spark 1.1 | unknown | $1.25 | $4.25 | 1,048,576 | 100.0% |
| Muse Spark 1.2 | unknown | $1.25 | $4.25 | 1,048,576 | 100.0% |
Everyone serving these models, cheapest first, with the numeric precision each one runs. The lab is often not the cheapest place to buy its own model.
| Model | Served by | Quant | $/M in | $/M out | Uptime 30m |
|---|---|---|---|---|---|
| Muse Glimmer | DeepInfra | bf16 | $0.30 | $1.20 | 100.0% |
| Muse Glimmer | Fireworks | unknown | $0.35 | $1.50 | 100.0% |
| Muse Glimmer | Phala | unknown | $0.35 | $1.50 | 97.2% |
| Muse Glimmer | Together | unknown | $0.35 | $1.50 | 99.7% |
| Muse Spark 1.1 | Meta | unknown | $1.25 | $4.25 | 100.0% |
| Muse Spark 1.2 | Meta | unknown | $1.25 | $4.25 | 100.0% |
| Llama 4 Maverick | DigitalOcean | unknown | $0.20 | $0.70 | 100.0% |
| Llama 4 Maverick | DeepInfra | fp8 | $0.20 | $0.80 | 99.6% |
| Llama 4 Maverick | Novita | fp8 | $0.27 | $0.85 | 99.0% |
| Llama 4 Maverick | Parasail | fp8 | $0.35 | $1.00 | 99.9% |
| Llama 4 Maverick | unknown | $0.35 | $1.15 | 98.0% | |
| Llama 4 Scout | DeepInfra | fp8 | $0.10 | $0.30 | 100.0% |
| Llama 4 Scout | Groq | unknown | $0.11 | $0.34 | 99.9% |
| Llama 4 Scout | Novita | bf16 | $0.18 | $0.59 | 99.9% |
| Llama 4 Scout | unknown | $0.25 | $0.70 | 99.7% |