Requesty
Fireworks AI

muse-glimmer-30b

Muse Glimmer 30B is a dense causal language model distilled from Muse Spark and purpose-built for autonomous agentic work. It combines multi-step reasoning, reliable schema-based tool calling, and failure recovery with multimodal understanding via a ~1.8B ViT-G/14 perception encoder, supporting interleaved text and image input, a 131K+ context window, and selectable reasoning strength (low through xhigh). Trained on data from over 100 languages, Muse Glimmer performs strongly for its size class on agentic benchmarks including MCP Atlas, DeepSearch QA, Gaia2 and SWE-Bench Pro, and is released under Apache 2.0.

VisionReasoningTool callingCachingJSON schema
CompareDocs

Specifications

Context window131K tokens
Max output131K tokens
API typechat
AddedAug 15, 2026
Model ID
Data retentionNo
Used for trainingNo
Provider location🇺🇸 US
Privacy policyFireworks Privacy Policy→

Benchmarks

Benchmarks haven't been published yet for this exact variant.

Some variants (region-specific deployments, highspeed tiers) share benchmarks with their base model. Check the base model page or the Fireworks AI models overview.

Pricing

Prices updated August 17, 2026
Input / 1M
$0.35
Output / 1M
$1.50
Cache write
N/A
Cache read / 1M
$0.04
Estimated cost
100K input + 10K output$0.0500
1M input + 100K output$0.50
10M input + 1M output$5.00

Requesty charges exactly what the upstream provider charges, no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to fireworks/muse-glimmer-30b.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="fireworks/muse-glimmer-30b",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other Fireworks AI models

qwen3.8-max
1M ctx$2.00/1M
Coding 72
nemotron-lightning-3.5-30b-a3b
262K ctx$0.05/1M
deepseek-v4-pro-0813
1M ctx$1.32/1M
Coding 69
deepseek-v4-flash-0731
1M ctx$0.14/1M
Coding 69
kimi-k3
1M ctx$3.00/1M
Coding 76
glm-5.2-fast
1M ctx$2.10/1M

Frequently asked questions

How much does muse-glimmer-30b cost?
muse-glimmer-30b is priced at $0.35 per million input tokens and $1.50 per million output tokens when accessed via Requesty. Prompt caching is supported, which can cut effective input cost by up to 90% on repeated context. Requesty charges exactly what the upstream provider charges, we don't add markup.
What is the context window of muse-glimmer-30b?
muse-glimmer-30b has a context window of 131K tokens, with a maximum output of 131K tokens per response. That's roughly 175 words of input you can fit in a single prompt.
What can muse-glimmer-30b do?
muse-glimmer-30b supports vision input, tool calling, extended reasoning, prompt caching, structured outputs (JSON schema). You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use muse-glimmer-30b with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "fireworks/muse-glimmer-30b". The Quickstart above shows Python, JavaScript and cURL snippets.
Can I run muse-glimmer-30b through Requesty?
Yes. muse-glimmer-30b runs through Requesty's OpenAI-compatible API, served from Fireworks AI. You do not host the model yourself: point base_url at Requesty, set the model to "fireworks/muse-glimmer-30b", and requests are routed to the upstream provider with automatic failover. The same key gives you 600+ other models too.

Access muse-glimmer-30b through Requesty

One API key, 600+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

All Fireworks AI models