Side-by-side comparison · models
Muse Glimmer 30B vs Muse Spark 1.1
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
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 1.1
Open Llama instruction model for multilingual chat, reasoning, and coding.
Checked facts
Line by line
| Measure | Muse Glimmer 30B | Muse Spark 1.1 |
|---|---|---|
| Input price | $0.300 per million tokens | $1.25 per million tokens |
| Output price | $1.20 per million tokens | $4.25 per million tokens |
| Price on deepinfra | $0.040 per million cached input tokens | Not checked yet |
| Price on deepinfra | $0.300 per million input tokens | Not checked yet |
| Price on deepinfra | $1.20 per million output tokens | Not checked yet |
| Price on together | $0.040 per million cached input tokens | Not checked yet |
| Price on together | $0.350 per million input tokens | Not checked yet |
| Price on together | $1.50 per million output tokens | Not checked yet |
| 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-1.1-20260709 |
| Time to first token (ms) on deepinfra | 443 | Not checked yet |
| Time to first token (ms) on together | 473 | Not checked yet |
| Output speed (tokens a second) on deepinfra | 110 | Not checked yet |
| Output speed (tokens a second) on together | 79 | Not checked yet |
| Time to first token (ms) | Not checked yet | 12,450 |
| Time to first token (ms) on meta | Not checked yet | 1,982 |
| Output speed (tokens a second) | Not checked yet | 258.5 |
| Output speed (tokens a second) on meta | Not checked yet | 203 |
| Tokens served on OpenRouter in a day | 2.2B (for 10 Nov 2026) | 551.2M (for 10 Nov 2026) |
| Tool calls on OpenRouter in a day | 50.2K (for 10 Nov 2026) | 6.6K (for 10 Nov 2026) |
| Artificial Analysis Intelligence Index | 6.3 | 33.7 |
| Artificial Analysis Coding Index | 49 | 71.3 |
| Artificial Analysis Agentic Index | 8.5 | 25.8 |
| Design Arena rating (Elo) | 1,200 | 1,281 |
| LMArena text rating (Elo) | Not checked yet | 1,479 |
| LMArena web development rating (Elo) | Not checked yet | 1,541 |
| LMArena vision rating (Elo) | Not checked yet | 1,292 |
| Simpleqa verified | Not checked yet | 57.8% (for 9 July 2026) |
| Apex agents | Not checked yet | 31.8% (for 9 July 2026) |
| Chess puzzles | Not checked yet | 34% (for 9 July 2026) |
| Dtbench | Not checked yet | 94.4% (for 9 July 2026) |
| Lmca | Not checked yet | 49.9% (for 9 July 2026) |
| Proofbench | Not checked yet | 39% (for 9 July 2026) |
| Surface evolver bench | Not checked yet | 52.5% (for 9 July 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.