Muse Spark Open (Xhigh)Ready
muse-spark-open
Serverless APILLMReasoningVisionPrompt cachingOpen weights
The compact, openly licensed member of the Muse family, tuned for high-throughput serving. It ranks #19 of the top 20 frontier models on the Artificial Analysis Intelligence Index with a score of 38.9. Released October 15, 2026.
1.0M
Context
816B
Parameters
#19
Intelligence rank · 38.9
Open
Weights · Muse Model License
View full spec
ReleasedOct 15, 2026
Knowledge cutoff—
Size classlarge
Input modalitiestext, image, speech, video
Output modalitiestext
Median output speed131 tok/s
Median time to first token0.7s
LicenseMuse Model License
Usage guide
1. Choose how you'd like to use Muse Spark Open (Xhigh)
☁️ Serverless
Use with pay-per-token pricing. Zero setup, no cold starts, automatic failover between upstream routes.
Usage tiers · Standard • Priority
Try it — sign in
curl https://tokenhub.coality.io/v1/chat/completions \
-H "Authorization: Bearer $TOKEN_HUB_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "muse-spark-open",
"messages": [{"role": "user", "content": "Explain vector databases in one paragraph"}]
}'Pricing
| Bucket | Price | Notes |
|---|---|---|
| Input tokens | $1.56 | per 1M tokens |
| Output tokens | $5.31 | per 1M tokens |
| Cache read | $0.1875 | cached input · 88% off |
| Cache write | — | per 1M tokens |
Billed per token with micro-dollar precision · effective Oct 27, 2026 · default key limits 600 req/min · 200K tokens/min
Benchmarks
GPQA Diamond
89.0
Humanity's Last Exam
43.9
MMMU-Pro
80.9
SciCode
58.4
Terminal-Bench 4.0
8.1
Long Context Reasoning
78.3
Scores from the Artificial Analysis evaluation suite (0–100 scale).
About this model
816B-parameter open-weight model, with a 1M context window, native image input, visible reasoning traces, prompt caching at a 88% discount on cached input, 131 output tokens/s median serving speed and 0.7s to first token.