Released August 9, 2026 — the first model from Meta Superintelligence Labs and Meta's return to a genuinely permissive licence. No Meta API price; hosted rates from third parties such as Together and OpenRouter land around $0.35/$1.50. Local hardware targets: 24GB for K-Quant-17GB, 32GB for K-Quant-Dynamic, 64GB for full precision.
76.0% on SWE-bench Verified — strong for a 30B dense model
Apache 2.0 licence with no restrictions on commercial use, modification or redistribution
Designed for long tool-call chains and failure recovery, with multimodal input and reasoning
Weaknesses
Qwen3.6-27B beats it on several practical agent and multimodal tests in Meta's own comparison table
Meta publishes no first-party API price — you pay a third-party host or run it yourself
131K context is small next to the 1M-token field
Real-world use cases
What people actually use Muse Glimmer 30B for.
Always-on local agents that make many sequential tool calls and must recover from failures
Commercial products that need unrestricted weights — Apache 2.0, no usage caps or redistribution limits
Running a capable agent model on a single 24GB or 32GB GPU via quantisation
How Muse Glimmer 30B compares
The nearest models people weigh against it, and what actually separates them.
vs Muse Spark — Against Muse Spark (Meta), Muse Glimmer 30B runs about 66% cheaper per token, gives up 8x on context and answers faster. Take Muse Glimmer 30B unless you specifically need what Muse Spark does better.
vs GPT-6 Astra — Against GPT-6 Astra (OpenAI), Muse Glimmer 30B runs about 97% cheaper per token, gives up 8x on context and answers faster. Take Muse Glimmer 30B unless you specifically need what GPT-6 Astra does better.
vs Claude Fable 5.1 — Against Claude Fable 5.1 (Anthropic), Muse Glimmer 30B runs about 97% cheaper per token, gives up 7.6x on context and answers faster. Take Muse Glimmer 30B unless you specifically need what Claude Fable 5.1 does better.
Price History
Muse Glimmer 30B pricing over time
→0% since Sep 1
7 data points · tracked daily since Sep 1, 2026
Ready to try it?
Start using Muse Glimmer 30B
Local and self-hosted agents that run continuously. Start free — no card required.
Meta Superintelligence Labs' first closed frontier model — a natively multimodal agentic reasoner (text, image, video, audio, PDF in) with a parallel-agent 'Contemplating mode', priced aggressively below rivals.
Verdict
Best-value multimodal agentic model — GPT-5.5-tier smarts, video/audio/PDF in.
Quality score
89%
Pricing
$1.25/1M in
$4.25/1M out
Speed
Balanced
3/5 speed
Context
1.0M tokens
v1.0 launched April 8, 2026 alongside Llama 5; v1.1 (July 9) opened the paid API; v1.2 (Aug 5) is coding-focused and powers Muse Code. Built with 'over an order of magnitude less' pretraining compute than Llama 4 Maverick. Cache hits $0.15/1M.
MultimodalAgenticValue1M context
Best for
Agentic tool-use and multimodal reasoning at aggressive pricing
OpenAI's September 3, 2026 frontier release — the first GPT-6 model and OpenAI's answer to Claude Fable 5.1 two days earlier. State of the art on computer use (OSWorld 2.0 72.6% in ~47% less time than GPT-5.6 Sol), agentic coding (Terminal-Bench 4.0 57.9%), and frontier math (FrontierMath Tier 4 97.6%). $10/$50 per 1M tokens, 1.05M context, 128K output, knowledge cutoff April 30, 2026.
Verdict
OpenAI's frontier answer to Fable 5.1 — computer-use and agentic-coding leader at $10/$50.
Quality score
99%
Pricing
$10.00/1M in
$50.00/1M out
Speed
Deliberate
2/5 speed
Context
1.1M tokens
Released September 3, 2026. API ID gpt-6-astra; rolling out over the coming days to ChatGPT Plus, Pro, Business and Enterprise (usage inside existing allowances; GPT-6 Astra Pro for Pro/Business/Enterprise; Enterprise off by default), the OpenAI API, Microsoft Azure and Amazon Bedrock. Standard API pricing $10/$50 per 1M tokens; Fast mode is up to 2x speed at 2x price; cache reads and writes have separate rates. Model docs list 1,050,000 context, 128,000 max output, knowledge cutoff April 30, 2026, reasoning efforts up to 'max'. Published launch numbers (Astra / GPT-5.6 Sol / Fable 5.1 / Opus 5): OSWorld 2.0 72.6 / 65.7 / — / 70.2; Terminal-Bench 4.0 57.9 / 37.3 / 55.8 / 52.3; Terminal-Bench Science 0.1 64.6 / 22.4 / 52.6 / 30.0; FrontierMath Tier 4 v2 97.6 / 83.0 / 87.8 / 73.2; GPQA Diamond 96.0 / 94.6 / 93.7 / 93.7; Humanity's Last Exam w/ tools 57.2 / — / 65.0 / 63.6; AutomationBench 41.4 / 18.1 / 31.4 / 26.9; DeepSWE v1.1 74.1 / 72.7 / 67.4 / 73.7; ARC-AGI-2 95.0 / 92.5 / 90.0 / 90.4; ARC-AGI-3 99.9 (OpenAI responses-API harness; ARC Prize's stateless runs score far lower) / 7.8 / — / 30.2; ExploitBench 100.0 / 78.5 / — / 70; SRE-Bench 88.0 / 55.9; Artificial Analysis Intelligence Index v4.1.1 61.2 / 60.9 / 65.7 / 63.1. Meets the Critical threshold for cybersecurity under OpenAI's Preparedness Framework; advanced cyber workflows gated behind OpenAI Daybreak. All figures from OpenAI's launch post and model docs, verified September 4, 2026.
Computer use leaderFrontierAgenticReasoningLong contextPremiumNew
Best for
Computer and browser use, long-horizon agentic coding, and frontier math and science work
Anthropic's September 1, 2026 frontier release and the new capability ceiling for coding, agents, and scientific work. Base pricing is unchanged at $10/$50, but cache reads dropped 75% to $0.25/1M — roughly 25% cheaper on typical workloads and up to 45% cheaper on agentic ones. 1M context, 128K output, adaptive thinking always on.
Verdict
New frontier leader — better than Fable 5 on every published benchmark, and cheaper to run.
Quality score
98%
Pricing
$10.00/1M in
$50.00/1M out
Speed
Deliberate
2/5 speed
Context
1M tokens
Released September 1, 2026 alongside Claude Mythos 5.1, the first update to the Mythos-class line since Fable 5 on June 9. API ID claude-fable-5-1; generally available on the Claude API, Amazon Bedrock, Google Cloud, and Microsoft Foundry. Published launch numbers (Fable 5.1 / Fable 5 / Opus 5 / GPT-5.6 Sol): Terminal-Bench-Science 0.1 52.6 / 24.7 / 29.0 / 22.4; Terminal-Bench 4.0 55.8 / 42.0 / 52.3 / 37.3; CursorBench 3.2.0 73.4 / 70.5 / 70.0 / 67.2; AutomationBench 31.4 / 17.1 / 26.9 / 19.6; OSWorld 2.0 strict 41.7 / 36.1 / 39.6; Humanity's Last Exam (no tools) 60.9 / 57.8 / 56.6; GDPval-AA v2 1853 / 1723 / 1824 / 1711. GDPval-AA v2 is rescaled from the v1 numbers quoted on the Fable 5 page and is not directly comparable to them.
Muse Glimmer 30B costs $0.35 per million input tokens and $1.5 per million output tokens on the API. A month of 10M input and 2M output tokens runs about $6.50 at list price, before any batch or caching discounts.
What is the context window of Muse Glimmer 30B?
Muse Glimmer 30B has a 131k tokens context window, with up to 131k tokens of output per response. That is the total of prompt plus response the model can hold in one request.
What is the knowledge cutoff of Muse Glimmer 30B?
Muse Glimmer 30B's training data runs through January 4, 2026, and the model was released on August 10, 2026. For anything after that date it needs web search or documents in the prompt.
What is Muse Glimmer 30B best for?
Muse Glimmer 30B is best for local and self-hosted agents that run continuously. It is a strong fit when that workflow matters more than the tradeoffs around budget pricing and fast speed.
When should I avoid Muse Glimmer 30B?
You need a large context window or the strongest agent scores at this size — check Qwen's 27B first.
What is a cheaper alternative to Muse Glimmer 30B?
Grok 4.5 (xAI) at $2.00/1M/1M input against Muse Glimmer 30B's $0.35/1M/1M. Best cost-per-solved-task coding agent — efficiency over ceiling. Compare it first if Muse Glimmer 30B's pricing is the thing stopping you.
What is a faster alternative to Muse Glimmer 30B?
Muse Spark — balanced against Muse Glimmer 30B's fast, with 1.0M tokens of context. Worth the swap when response time is what your users notice rather than the last few points of reasoning depth.
Newsletter
Get notified when Muse Glimmer 30B pricing changes
We track pricing daily. When this model drops or spikes, you'll know first.
No spam. Useful updates only. Affiliate disclosures always clearly labeled.