Muse Glimmer 30B vs Muse Spark 1.1
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
| Specification |
by Fireworks
2 providers:
Fireworks $0.350/$1.50
Together $0.350/$1.50
|
by Meta
|
|---|---|---|
| Overview | ||
| Status | Current budget Open weights | Current flagship |
| Released | Aug 9, 2026 | Jul 9, 2026 |
| Pricing per million tokens | ||
| Input | $0.350/Mtok | $1.25/Mtok |
| Output | $1.50/Mtok | $4.25/Mtok |
| Blended avg | $0.637/Mtok | $2.00/Mtok |
| Cached input | $0.040/Mtok | $0.150/Mtok |
| Price basis | — | Meta Model API list price, per 1M tokens, with no long-context premium. Search grounding is billed separately at $2.50 per 1,000 queries and is not included in the token rate. Public preview was US-only at launch. |
| Specifications | ||
| Context window | 131K tokens | 1M tokens |
| Parameters | 30B | Proprietary |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
text
|
textimagevideoaudio
|
| Benchmarks sources: Artificial Analysis / Artificial Analysis, LMSYS Chatbot Arena (UC Berkeley), Meta evaluation report | ||
| Independent composite scores measured for both models — each on its own scale | ||
| AA Intelligence Index | 35.1 | 53.2 |
| AA Agentic Index | 22.9 | 39.7 |
| AA-Omniscience Index (−100–100) | -32.9 | 28.1 |
| GDPval-AA v2 | — | 1370.6 ELO |
| LMSYS Chatbot Arena | — | 1490.4 ELO |
| Per-benchmark results published for Muse Spark 1.1 only — no independent per-benchmark scores exist for Muse Glimmer 30B yet | ||
| Avg benchmark score | — | 61.5 |
| Perf / dollar | — | 30.8 |
| Terminal-Bench 2.1 | — | 78 |
| MCP Atlas | — | 88.1 vendor |
| OSWorld-Verified | — | 80.8 vendor |
| Humanity's Last Exam | — | 45 |
| Providers | ||
| Available from |
Fireworks — $0.350/$1.50/Mtok
Together — $0.350/$1.50/Mtok
|
Meta — $1.25/$4.25/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | Muse Glimmer 30B | Muse Spark 1.1 | Savings |
|---|---|---|---|
| 1M tokens | $0.64 | $2 | $1.36 (68%) |
| 10M tokens | $6.38 | $20 | $13.63 (68.2%) |
| 100M tokens | $63.75 | $200 | $136.25 (68.1%) |
| 1000M tokens | $637.5 | $2000 | $1362.5 (68.1%) |
When to pick which
distilled from the pricing and spec data above- cost dominates: $0.637/Mtok blended vs $2.00 — 68.1% less on the same 50/50 token mix
- your workload re-reads context (agents, RAG, long chats): cached input costs $0.040/Mtok — 89% off its list input price
- you want open weights — self-host it, fine-tune it, or exit the API entirely; Muse Spark 1.1 is closed
- provider flexibility: available from 2 providers, so you can price-shop hosts or fail over
- you need the longer context: 1M tokens vs 131K (8×)
The cheaper option here — Muse Glimmer 30B costs $0.637/Mtok blended on Fireworks.
Summary
Muse Glimmer 30B by Fireworks costs $0.350/Mtok input and $1.50/Mtok output, with a 131K-token context window. It supports text input and is available from 2 providers.
Muse Spark 1.1 by Meta costs $1.25/Mtok input and $4.25/Mtok output, with a 1M-token context window. It supports text, image, video, audio input.
On a blended cost basis, Muse Glimmer 30B is 68.1% cheaper than Muse Spark 1.1.
The two aren't directly comparable on average benchmark score: Muse Spark 1.1 has published per-benchmark results, while Muse Glimmer 30B does not yet — it is measured today on independent composites (see the table above).
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.