Side-by-side comparison · models
Muse Glimmer 30B (batch) vs Muse Spark 1.3
What we have checked about each, line by line. Every value shows where it came from and when we last confirmed it.
Models
Muse Glimmer 30B (batch)
Muse Glimmer 30B is a dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware.
Models
Muse Spark 1.3
Muse Spark 1.3 is a multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows.
Checked facts
Line by line
| Measure | Muse Glimmer 30B (batch) | Muse Spark 1.3 |
|---|---|---|
| Input price | Not checked yet | $1.25 per million tokens |
| Output price | Not checked yet | $4.25 per million tokens |
| Price on meta | Not checked yet | $0.150 per million cached input tokens |
| Price on meta | Not checked yet | $1.25 per million input tokens |
| Price on meta | Not checked yet | $4.25 per million output tokens |
| Latest version | meta/muse-glimmer-30b-20260810 | meta/muse-spark-1.3-20260902 |
| Time to first token (ms) | Not checked yet | 36,130 |
| Time to first token (ms) on meta | Not checked yet | 3,605 |
| Output speed (tokens a second) | Not checked yet | 140.2 |
| Output speed (tokens a second) on meta | Not checked yet | 95 |
| Tokens served on OpenRouter in a day | Not checked yet | 7.4B (for 7 Nov 2026) |
| Tool calls on OpenRouter in a day | Not checked yet | 59.8K (for 7 Nov 2026) |
| Artificial Analysis Intelligence Index | Not checked yet | 36.6 |
| Artificial Analysis Coding Index | Not checked yet | 76.5 |
| Artificial Analysis Agentic Index | Not checked yet | 55.5 |
| LMArena text rating (Elo) | Not checked yet | 1,491 |
| LMArena web development rating (Elo) | Not checked yet | 1,654 |
| LMArena vision rating (Elo) | Not checked yet | 1,308 |
| Design Arena rating (Elo) | Not checked yet | 1,390 |
| Frontiermath tiers 1 3 v2 private | Not checked yet | 74.4% (for 2 Sept 2026) |
| Apex agents | Not checked yet | 58.6% (for 2 Sept 2026) |
| Chess puzzles | Not checked yet | 37% (for 2 Sept 2026) |
| Dtbench | Not checked yet | 96.5% (for 2 Sept 2026) |
| Frontiermath tier 4 v2 private | Not checked yet | 46.3% (for 2 Sept 2026) |
| Frontierswe | Not checked yet | 25.8% (for 2 Sept 2026) |
| Lmca | Not checked yet | 53.9% (for 2 Sept 2026) |
| Mystery game puzzles | Not checked yet | 25% (for 2 Sept 2026) |
| Proofbench | Not checked yet | 58% (for 2 Sept 2026) |
| Weirdml v3 | Not checked yet | 12.1% (for 2 Sept 2026) |
| Otis mock aime 2024 2025 | Not checked yet | 99.2% (saturated benchmark: the top models all score near the ceiling) (for 2 Sept 2026) |
A line appears only when at least one record has a value we confirmed from a source. Values we have not confirmed are left out, never guessed. Records of different types share fewer lines.
Add a third
Compare with one more record
How this works
Share it, or build your own
This page works for any two or three records in the directory, of any type. The link stays the same whichever order you name them in, so you can share it as it is. Change the records in the address, for example /directories/vs/claude-code-vs-cursor, or start from our editorial comparisons.