Models

Meta ships six model lines. Muse Wonder, Muse Spark, Muse Image, Muse Voice Transcribe, and Segment Anything Model (SAM) are hosted on Meta Model API; Muse Glimmer is open-weight and runs on your own hardware. Muse Spark also has an open-weight release, Muse Spark Open. Pick one by what you're building, then read its section for versions and where to use them.

Models at a glance

Muse Wonder

Muse Wonder is Meta's most capable model, built for complex software engineering and long-horizon agentic work. It makes and verifies changes across many files in real repositories, and it plans, delegates to sub-agents, calls tools in parallel, and recovers from errors across many steps, following detailed instructions closely. Take on hard coding tasks, build personal and professional agents, and drive a desktop with computer use. Muse Wonder also handles chat completion, image understanding, video and audio understanding, tool calling, structured output, and search grounding.

Muse Wonder takes text, image, video, audio, and PDF input, returns text, has a 1,048,576-token context window, and supports reasoning effort levels from "minimal" to "max".

Muse Wonder availability

Muse Wonder comes in two variants. Muse Wonder is on Model API and in Muse Code. Muse Wonder Full, the higher-capability variant, is available only in Muse Code with a subscription, so setting reasoning effort to "max" on Muse Wonder doesn't select it.

With a subscription, Muse Wonder is available on every plan and Muse Wonder Full on the Power Usage and Maximum Usage plans, each in Standard mode and Contributor mode.

In Standard mode, your content — the prompts, code, files, and context you send, plus the model responses you get back — is not used to train Meta models. Muse Code subscribers can also use Contributor mode, which is on by default: it gives you more usage in exchange for letting Meta use your content to train future Meta models, and you can turn it off at any time in your settings. For Muse Wonder, Contributor mode isn't available on pay-as-you-go. See Pricing and rate limits for rates.

Muse Spark

Muse Spark is Meta's model for agentic and coding work — multi-step tool loops, software engineering assistants, and long-context reasoning. Use it for chat completion, image understanding, video and audio understanding, tool calling, structured output, and search grounding.

Muse Spark comes in three versions. All take text, image, video, audio, and PDF input, return text, and have a 1,048,576-token context window. They differ by capability:

• Muse Spark 1.3 (muse-spark-1.3): the latest version, tuned for agentic workflows (multi-step tool, browser, and long-horizon tasks) with improved coding over 1.2. Supports all reasoning effort levels, including the "max" level for extended reasoning (not available with muse-spark-1.3-contributor). Recommended for new work.
• Muse Spark 1.2 (muse-spark-1.2): the previous version.
• Muse Spark 1.1 (muse-spark-1.1): the original version.

Audio understanding in Muse Spark 1.3 is currently not fully supported, and response quality for requests including audio content may be degraded. For audio, use Muse Spark 1.2, or Muse Voice Transcribe for dedicated speech-to-text.

muse-spark-1.3 is the default model in most code examples in these docs. The coding agents, agent frameworks, and computer use guides use muse-wonder-1.0 for long-running agentic work.

Muse Spark availability

Every Muse Spark version is available on pay-as-you-go. All versions except Muse Spark 1.1 are also available in Muse Code, on pay-as-you-go and with any subscription.

In Standard mode, your content is not used to train Meta models. In Contributor mode, your content may be used to train future Meta models, in exchange for a lower price. On pay-as-you-go, switch between the two by choosing the model ID: for example, muse-spark-1.3 for Standard mode or muse-spark-1.3-contributor for Contributor mode. Muse Spark 1.1 is available in Standard mode only. See Pricing and rate limits for rates.

Choose Muse Wonder or Muse Spark

Muse Wonder and Muse Spark take the same requests on the same endpoints, so moving a workload between them is a model ID change. Choose by workload:

• Muse Wonder: long-running, complex tasks where capability matters more than cost per token, such as multi-file coding changes, multi-hour agent runs, and sub-agent orchestration.
• Muse Spark: high-volume, latency-sensitive, or cost-sensitive work, such as sub-agent tasks, classification, and tight tool loops. On Model API, Muse Spark also offers a discounted Contributor mode.

Muse Image

Muse Image is Meta's image generation model line. Send a text prompt (and optional reference images) and get an image back, or send an image with an instruction to edit it. One model handles both generation and editing. It's agentic, so it can search the web for visual references and current facts and run code to build layouts before it renders.

Reach Muse Image two ways: the Responses API for conversational, multi-turn editing, or the single-shot /v1/images/generations and /v1/images/edits endpoints for one-off calls. See the Image generation guide to get started.

Muse Voice Transcribe

Muse Voice Transcribe is Meta's speech-to-text model on Meta Model API. Use it to transcribe live audio streams or supported audio files.

Muse Voice Transcribe is built for real-time speech products: voice agents, meeting and call intelligence, live transcription, dictation, captioning, and high-volume transcription. It supports streaming speaker diarization, native endpointing and voice activity detection, contextual and keyword biasing, 25 evaluated languages with code-switching, and turn-level timestamps.

It returns transcript text. It does not synthesize speech or provide a speech-to-speech conversation API. It does not provide word-level timestamps, sound event detection, or emotion detection.

Reach Muse Voice Transcribe two ways: the realtime WebSocket endpoint (wss://api.meta.ai/v1/asr/realtime) for live audio, or the file endpoint (POST /v1/asr/transcribe) for a recording you already have.

Start with the Muse Voice Transcribe guide. See Pricing and rate limits for audio pricing.

Segment Anything Model

SAM 3.1 is Meta's segmentation model on Meta Model API. Give it a short noun phrase such as "yellow school bus" with an image or video, and it returns every match as a box and a pixel-accurate mask. One model handles both stills and clips, and in video it follows each object across frames.

Reach SAM 3.1 over the Responses API (POST /v1/responses), and upload clips through the Files API when you segment them more than once. See the Media segmentation guide to get started.

Muse Spark Open

Muse Spark Open is the open-weight release of Muse Spark: an 816B-parameter mixture-of-experts model with 42B active parameters. You don't call it over Model API. Download the weights and serve them through a runtime such as SGLang, vLLM, TensorRT-LLM, or llama.cpp, or call it through a hosted inference provider.

Muse Spark Open lives in the Open-weight models section:

• Muse Spark Open overview: checkpoints, hardware, license, and launch partners.
• Get the model: download the BF16 or NVFP4 weights.
• Run inference: serve it on your own GPUs or call a hosted provider.
• Post-training: adapt it with supervised fine-tuning (SFT) or reinforcement learning (RL).

Muse Glimmer

Muse Glimmer is Meta's open-weight multimodal model, distilled from Muse Spark and built to run on your own hardware. Unlike Muse Wonder, Muse Spark, Muse Image, Muse Voice Transcribe, and SAM, you don't call it over Model API — you download the weights and serve it through a runtime such as vLLM, SGLang, llama.cpp, or ExecuTorch.

Muse Glimmer lives in the Open-weight models section, covering how to get, prompt, deploy, and customize the model:

• Muse Glimmer overview: variants, architecture, license, and launch partners.
• Get the model: download the weights and verify your setup.
• Run inference: pick a runtime and serve it locally.
• Post-training: adapt it to your domain with supervised fine-tuning (SFT) or reinforcement learning (RL).

List models via the API

Query the catalog programmatically when you need to check what's enabled for your team. This returns the API-hosted models; self-hosted Muse Glimmer isn't included.

python
import os
​
from openai import OpenAI
​
client = OpenAI(
    base_url="https://api.meta.ai/v1",
    api_key=os.environ["MODEL_API_KEY"],
)
​
response = client.models.list()
​
print(response.model_dump_json(indent=2))

python
import json
import os
​
import requests
​
response = requests.get(
    "https://api.meta.ai/v1/models",
    headers={"Authorization": f"Bearer {os.environ['MODEL_API_KEY']}"},
)
response.raise_for_status()
print(json.dumps(response.json(), indent=2))

shell
curl -X GET "https://api.meta.ai/v1/models" \
  -H "Authorization: Bearer $MODEL_API_KEY"

The response returns models sorted newest-first by creation time, with the model ID as a stable tiebreaker. Each object includes a created field — the Unix timestamp in seconds for when the model was added to the registry. GET /v1/models/{model} returns the same created value for a single model.

See also

• Pricing and rate limits: per-token rates for Standard and Contributor mode, and rate limits for each model
• Subscriptions: Muse Code plans and what each one includes