Muse Wonder
muse-wonder- 🏢 Frontier reasoning model from Meta, released October 2026
- 📏 Roughly 1,048,576-token context window; text output
- 👁️ Accepts text, images, video, audio and PDF documents
- 🔧 Plans, delegates to sub-agents, calls tools in parallel
- 🧠 Recovers from errors so long-horizon jobs finish unattended
- 🎯 Built for complex personal and professional agentic tasks
- 💬 Function calling, web search and code-oriented work supported
- 🆕 Released days before Meta's Muse Spark Open in this catalog
Meta Platforms is an American multinational technology company headquartered in Menlo Park, California. Best known for operating Facebook, Instagram, WhatsApp, and Threads, Meta has become one of the most influential players in open-weight AI research through its Llama family of…
Explore 6 more models by Meta Platforms →Muse Wonder is Meta's frontier reasoning model for complex personal and professional agentic tasks, listed with a release date of October 12, 2026 and documented on Meta's developer site. It is positioned for long-horizon work: rather than answering a single prompt, it is meant to carry a multi-step job — research, paperwork, software building — through to completion without step-by-step direction.
The agentic design is the defining feature. The model plans a task, delegates portions to sub-agents, calls tools in parallel rather than strictly in sequence, and recovers from errors mid-run so a workflow can continue instead of stalling. Capabilities registered for it include reasoning, vision, code-oriented generation, function calling, web search and audio input.
Within Meta's current lineup, Muse Wonder sits alongside Muse Spark Open, an open-weights text model released days later, and the image-generation and editing pair Muse Image. It represents a different track from Meta's earlier openly distributed Llama models in this catalog, such as Llama 3.3 70B and Llama 3.2 3B, which are smaller and shipped as downloadable weights.
On inputs and limits: text, images, video, audio and PDFs go in, text comes out, and the context window is about one million tokens.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usage · Muse Wonder · Organic sellers
Top sellers · Muse Wonder · Organic
- ▲ Apex Ant68.8%
- Super Seeder17.1%
- ZLKPro-Api10.8%
- Fire Ant 🔥🐜3.3%
- surplus-hive0.1%
- OpenCheap0.0%
- Others0.0%
| Seller | Health | Rep↓ | Sales | Tokens | Routing | Input $/M | Cached $/M | Output $/M | Save |
|---|---|---|---|---|---|---|---|---|---|
| Super Seeder 0xd19f…41f3 | healthy | 89.22 | $4.53 | 1.73M | #2 | $6.2501 | $0.3125 | $31.2506 | — |
| ▲ Apex Ant 0x73b4…e736 | healthy | 81.71 | $18.27 | 23.02M | #1 | $2.50 | $0.13 | $12.50 | −38% |
| OpenCheap 0x0ee1…6c41 | — | 12.43 | $0.01 | 4.55k | gated | $2.0825 | $2.061 | $10.4125 | −48% |
| Sapiens 0x1fae…8318 | healthy | 8.21 | $0.00 | 175.00 | gated | $2.125 | $2.1031 | $10.625 | −47% |
| dealer 0xc46e…8ce6 | — | 2.57 | — | — | gated | $2.4719 | $0.1285 | $12.3594 | −38% |
| Fire Ant 🔥🐜 0xbe05…bc5d | — | 0.00 | $0.87 | 258.97k | gated | $4.50 | $0.225 | $22.50 | — |
"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.