Checked: Muse Glimmer 30B OpenRouter price vs HF downloads
On 2026-11-02, Muse Glimmer 30B lists at $0.30 in and $1.20 out per million tokens, context 131,072. Official downloads 195,811; Unsloth GGUF 812,668.
You are checking the OpenRouter list price and context for Muse Glimmer 30B, and the Hugging Face download counts on the official repo and the Unsloth GGUF repo. The figures below were read on 2026-11-02 from the public models list and the Hub search API.
The OpenRouter id is meta/muse-glimmer-30b (display name Meta: Muse Glimmer 30B). created is 1786302394, that is 2026-08-09 19:06 UTC. The official repo was created on 2026-08-09, and the Unsloth GGUF repo on 2026-08-10. This page lines up OpenRouter list price and context with those two Hub download counts. A download count is a repo counter, not a quality score.
List price and context on OpenRouter
context_length is 131,072. pricing.prompt is 0.0000003 USD per token and pricing.completion is 0.0000012 USD per token. Times 1,000,000, input is $0.30 and output is $1.20. input_cache_read is 0.00000004, or $0.04 per million tokens. max_completion_tokens is 16,384, shorter than the full context. architecture is text+image→text. is_moderated is false. canonical_slug is meta/muse-glimmer-30b-20260810. Both Hub repos below list pipeline_tag image-text-to-text. The API description calls it a dense open-weight multimodal model distilled from Muse Spark for agents on consumer hardware. That sentence has no score, so the table leaves it out.
| Source | Identifier | Created | Figure on 2026-11-02 |
|---|---|---|---|
| OpenRouter API | meta/muse-glimmer-30b | 2026-08-09 | 131,072 context · $0.30 in · $1.20 out / 1,000,000 tokens |
| Hugging Face official weights | meta-models/Muse-Glimmer-30B | 2026-08-09 | 195,811 downloads · 2,108 likes |
| Hugging Face GGUF (Unsloth) | unsloth/Muse-Glimmer-30B-GGUF | 2026-08-10 | 812,668 downloads · 588 likes |
On the snapshot day 2026-11-02, the official repo had been public for 85 days, with 195,811 downloads and 2,108 likes. The Unsloth GGUF repo had been public for 84 days, with 812,668 downloads and 588 likes. 812,668 ÷ 195,811 ≈ 4.15, which is a ratio of download counts. Likes rank the repos the other way. The same search also lists the official GGUF repo meta-models/Muse-Glimmer-30B-GGUF at 244,803 downloads and 388 likes, createdAt 2026-08-09. This table does not add those counters together. Downloads and likes are Hub counters, not a quality score.
How to re-read the two catalogs
- 1
Find the OpenRouter row
Open https://openrouter.ai/api/v1/models and locate id
meta/muse-glimmer-30b. Copy context_length, pricing.prompt, pricing.completion, pricing.input_cache_read, and created. - 2
Scale USD per token to one million
0.0000003 times 1,000,000 is $0.30 per million tokens. 0.0000012 becomes $1.20. 0.00000004 becomes $0.04 for a cache read.
- 3
Read two Hub repos separately
Open https://huggingface.co/api/models?search=Muse-Glimmer-30B . Record downloads, likes, and createdAt for meta-models/Muse-Glimmer-30B and for unsloth/Muse-Glimmer-30B-GGUF. The Unsloth repo is a community upload. Do not add meta-models/Muse-Glimmer-30B-GGUF into the official repo’s download count.
- 4
Read likes beside downloads
The official repo shows 195,811 downloads and 2,108 likes. The Unsloth GGUF repo shows 812,668 downloads and 588 likes. The two counters rank the repos differently. Both are Hub counters, not an eval score.
What each figure is for
The list price is the hosted catalog rate
$0.30 / $1.20 per million tokens comes from OpenRouter pricing. These two Hub rows do not list an inference price.
Context and the completion cap differ
context_length is 131,072. max_completion_tokens is 16,384.
Downloads and likes rank differently
The official repo shows 195,811 downloads and 2,108 likes. The Unsloth GGUF repo shows 812,668 downloads and 588 likes. A download count is not a quality score.
To spread these figures on one shared page, open a browser board at wbstorm. Room setup is in the guide. A related catalog check is Muse Spark Open price and downloads.
Fields copied from the two APIs
# fetched 2026-11-02
# https://openrouter.ai/api/v1/models
# id: meta/muse-glimmer-30b
# name: Meta: Muse Glimmer 30B
# created: 1786302394 = 2026-08-09T19:06:34Z
# canonical_slug: meta/muse-glimmer-30b-20260810
# context_length: 131072
# pricing.prompt: 0.0000003 USD/token = 0.30 USD / 1000000 tokens
# pricing.completion: 0.0000012 USD/token = 1.20 USD / 1000000 tokens
# pricing.input_cache_read: 0.00000004 USD/token = 0.04 USD / 1000000 tokens
# max_completion_tokens: 16384
# is_moderated: false
# modality: text+image->text
# https://huggingface.co/api/models?search=Muse-Glimmer-30B
# meta-models/Muse-Glimmer-30B downloads=195811 likes=2108 createdAt=2026-08-09T17:51:35.000Z
# unsloth/Muse-Glimmer-30B-GGUF downloads=812668 likes=588 createdAt=2026-08-10T09:11:29.000Z
# meta-models/Muse-Glimmer-30B-GGUF downloads=244803 likes=388 createdAt=2026-08-09T17:54:22.000ZWhich figures does this page line up?
The OpenRouter price per million tokens, the 131,072 context length, and the download and like counts on the official repo and the Unsloth GGUF repo, as fetched on 2026-11-02.
How do you turn the API price into a per-million figure?
pricing.prompt and pricing.completion are USD per token. Multiply by 1,000,000. 0.0000003 becomes $0.30 and 0.0000012 becomes $1.20. A cache read at 0.00000004 becomes $0.04 per million tokens.
Is a download count a quality score?
A download count is the Hugging Face downloads field, not a quality score. The Unsloth GGUF repo has the higher download count. The official repo has the higher like count. The two counters rank the repos differently, and neither is an eval score.
Were the official repo and the Unsloth GGUF created on the same day?
The official createdAt is 2026-08-09 17:51 UTC. OpenRouter created is unix 1786302394, which is 2026-08-09 19:06 UTC. The Unsloth GGUF createdAt is 2026-08-10 09:11 UTC.