Compare / Muse Glimmer 30B vs Gemma 4 31B
Muse Glimmer 30B vs Gemma 4 31B
Updated November 2026
Verdict
Muse Glimmer 30B scores higher on the Small LM Index: 72.6 vs 70.1, although their likely ranges overlap, so the lead is not certain. On the 6 benchmarks both labs report, Muse Glimmer 30B wins 4 and Gemma 4 31B wins 2. Muse Glimmer 30B is smaller: it needs about 19.2 GB of memory at 4-bit, against 21.9 GB for Gemma 4 31B. Gemma 4 31B is cheaper through an API: $0.16 per million tokens against $0.53.
Pick Muse Glimmer 30B if…
- you want the strongest model overall
- you mostly write code
- you need math and reasoning
- you build agents that call tools
- you have limited memory (about 19.2 GB at 4-bit)
Pick Gemma 4 31B if…
- API cost matters most
- you work with long documents
- you want the more widely used model (more quantized builds and community support)
Specifications
| Muse Glimmer 30B | Gemma 4 31B | |
|---|---|---|
| Small LM Index | 72.6 (likely 71–74) | 70.1 (likely 69–72) |
| Lab | Meta · United States | Google · United States |
| Released | 2026-08-09 | 2026-04-02 |
| Parameters | 29.6B (dense) | 30.7B (dense) |
| Memory at 4-bit | ~19.2 GB | ~21.9 GB |
| Context window | 131K tokens | 256K tokens |
| Input | text, image | text, image |
| Reasoning | Always thinks | Hybrid (thinking on/off) |
| License | Apache-2.0 | Apache-2.0 |
| API price / 1M tokens | $0.53 ($0.30 in, $1.20 out) | $0.16 ($0.10 in, $0.34 out) |
| Hugging Face downloads (30 days) | 189.7K | 8M |
Benchmarks
Higher is better. Bold marks the winner where both labs report a score. ≈ marks the index model's estimate for a score that wasn't reported, ✓ an independent measurement and * a score measured by another lab.
| Benchmark | Muse Glimmer 30B | Gemma 4 31B |
|---|---|---|
| MMLU-Pro Knowledge | ≈ 86 | 85.2 |
| GPQA Diamond Knowledge | 83.5✓ | 84.3 |
| Humanity's Last Exam Reasoning | 22.0✓ | 19.5 |
| AIME Math | 94.7 | 89.2 |
| LiveCodeBench Coding | ≈ 83 | 80.0 |
| SWE-bench Verified Coding | 76.0 | 52.0* |
| Terminal-Bench Agentic | 51.7 | 42.9* |
| τ²-Bench Agentic | — | 76.9 |
| IFBench Instruction following | 77.0 | — |
| MMMU-Pro Vision | 74.0 | 76.9 |
Scores are mostly reported by each lab with its own settings, so small gaps are noise. See the methodology. Open in the interactive comparison to add more models.
Questions
Which is better, Muse Glimmer 30B or Gemma 4 31B?
Muse Glimmer 30B scores higher on the Small LM Index: 72.6 vs 70.1, although their likely ranges overlap, so the lead is not certain. On the 6 benchmarks both labs report, Muse Glimmer 30B wins 4 and Gemma 4 31B wins 2.
Which is cheaper to run?
Gemma 4 31B is cheaper through an API: $0.16 per million tokens against $0.53. Muse Glimmer 30B is smaller: it needs about 19.2 GB of memory at 4-bit, against 21.9 GB for Gemma 4 31B.
Which has the longer context window?
Gemma 4 31B has the longer context window: 256K tokens against 131K.
Can I use Muse Glimmer 30B and Gemma 4 31B commercially?
Both are released under the Apache-2.0 license. Apache 2.0 and MIT allow commercial use; other licenses can add conditions, so read the license before shipping a product.