Muse Glimmer 30B
Last verified August 11, 2026 · Meta pricing ↗$0.350/1M input · $1.50/1M output · 131K context · Meta
Muse Glimmer 30B costs $0.350 per 1M input tokens and $1.50 per 1M output tokens from Meta, with a 131K-token context window. A typical chat reply (1,500 tokens in, 400 out) costs $0.0011, or $11.25 per 10,000 requests. On that workload it is cheaper than 118 of the 171 other buyable models tracked here.
Count Tokens
—
Tokens
—
Input Cost
—
Output Cost
Estimate Monthly Cost
Monthly Cost Estimator
Quick:
<$0.0001/mo
Pick a preset above or enter custom usage
What Muse Glimmer 30B Costs Per Request
At $0.350 per 1M input and $1.50 per 1M output tokens, before caching or batch discounts.
| Workload | Input / output tokens | Per request | Per 10,000 requests |
|---|---|---|---|
| Chat reply | 1,500 / 400 | $0.0011 | $11.25 |
| RAG answer over retrieved docs | 8,000 / 600 | $0.0037 | $37.00 |
| Coding-agent step | 30,000 / 2,000 | $0.013 | $135.00 |
Muse Glimmer 30B vs Similarly Priced Models
Nearest buyable models by price, with the RAG answer above (8,000 in / 600 out) priced on each.
| Model | Input /1M | Output /1M | Context | RAG answer |
|---|---|---|---|---|
| Muse Glimmer 30B Meta | $0.350 | $1.50 | 131K | $0.0037 |
| Llama 4 Maverick Meta | $0.270 | $0.850 | 1.0M | $0.0027 |
| Gemini 3.1 Flash-Lite Google | $0.250 | $1.50 | 1.0M | $0.0029 |
| GPT-4.1 Mini OpenAI | $0.400 | $1.60 | 1.0M | $0.0042 |
Alternatives to Muse Glimmer 30B
Pricing Details
MetaMeta's first open weights in more than a year, Apache 2.0, ~29.6B dense including a ~1.8B vision encoder (not MoE), logit-distilled from Muse Spark during pre-training (Meta names no version). Text and image in, text out, 100+ languages, knowledge cutoff January 4, 2026. Meta does not serve it on its own API and points developers to Hugging Face. The $0.35/$1.50 rate is Together AI's, and as of August 11, 2026 Together is the only host publishing a per-token price: OpenRouter's endpoints API returns exactly one endpoint (provider Together) at the identical figures, ModelsLab resells at the same card, NVIDIA build.nvidia.com offers a rate-capped free prototyping tier with no per-token rate, and Fireworks was named a launch partner but has shipped no listing. So the cross-host spread is 1.0x, against 11.7x for gpt-oss-120b and 12.4x for Gemma 4 31B. Max output tokens are published nowhere (Together returns null); 16384 here is a catalogue default, not a documented limit. Artificial Analysis' model page lists a 256K context, contradicted by the config, Together and NVIDIA NIM, which all say 131,072; the 32,768 figure in Meta's launch blog sits inside an example agent config. Benchmarks below are Meta-reported. On Meta's own nine-row table it wins five and loses four against Qwen3.6-27B; the 60.7 TerminalBench 2.1 figure widely attributed to Muse Glimmer is Qwen's score, and Meta's card says 51.7. SWE-bench Verified 76.0, one of the rows it loses: Meta's own table puts Qwen3.6-27B in thinking mode at 77.2 on the same suite (both figures read off Meta's model page 2026-08-11). Independent AA figures agree there (52%) but are lower overall: Intelligence Index 35 against Qwen3.6 27B's 38, GDPval-AA v2 Elo 953 against 1141, and an 82% hallucination rate against Qwen's 49%. Openness Index 44. 4-bit GGUF is 15.9 GB and runs on a 24GB card; BF16 shards total 55.7 GB. Estimated tokens (different tokenizer).
Input / 1M tokens
$0.35
Output / 1M tokens
$1.5
Context Window
131K
Max Output
16K
Price History
No price change recorded since 2026-08-11, when this site started tracking this model; it launched on 2026-08-10.
Frequently Asked Questions
Common questions about Muse Glimmer 30B pricing and usage
How much does Muse Glimmer 30B cost?
Muse Glimmer 30B costs $0.35 per 1 million input tokens and $1.5 per 1 million output tokens. This pricing is set by Meta.
What tokenizer does Muse Glimmer 30B use?
Muse Glimmer 30B uses the cl100k_base tokenizer. This tokenizer has a vocabulary of ~100,000 tokens and is widely used across many models.
What is Muse Glimmer 30B's context window?
Muse Glimmer 30B supports a context window of 131K tokens (131,072 tokens). The maximum output length is 16K tokens.
How does Muse Glimmer 30B compare to other models?
On an 8,000-token-in, 600-token-out RAG answer, Muse Glimmer 30B costs $0.0037 per request. The nearest-priced alternatives cost: Llama 4 Maverick (Meta) $0.0027, Gemini 3.1 Flash-Lite (Google) $0.0029, GPT-4.1 Mini (OpenAI) $0.0042. The table above lists their full rates and context windows.
How do I count tokens for Muse Glimmer 30B?
Use the token counter at the top of this page or visit the homepage calculator. Paste your text and it will show the exact token count and cost for Muse Glimmer 30B using the cl100k_base tokenizer.