Head to head · LLM inference · October 2026 research run

Muse Ember vs Muse Spark Open

Muse Spark Open scores 53.4 (D) on agent readiness against Muse Ember's 51.8 (D), and leads in 4 of 7 scored categories. Muse Ember leads on security & auth. Both do llm inference.

Best model APIs and inference for AI agents · All 871 models comparisons

Which one, for what

Muse Ember D

Best for Agents that hand out many small, bounded tasks (sub-agents, tool loops, routing, extraction over thousands of files) and want a cheap multimodal model with a 1M context behind the same API as a bigger Muse model.

Ahead on

  • Security & auth, 64 against 44

Also in its favour

  • A hosted endpoint, with nothing to install

Watch for

No weights and no parameter count published, so it can't be run locally

Muse Spark Open D

Best for A team with a four-GPU Blackwell node, or two to four Mac Studios, that wants a 1M-token multimodal coding and agent model on its own hardware with no prompts leaving it.

Ahead on

  • Payments & pricing, 50 against 20
  • Maintenance & community, 60 against 46

Also in its favour

  • No key needed to call it

Watch for

The NVFP4 checkpoint (about 537 GB) serves on four B300 GPUs and the BF16 one (about 1.65 TB) on eight

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Payments & pricing Muse Spark Open +30

Muse Ember 20

Published rubric. No payment protocol (0). Per-token prices for both tiers published without a login, read 9 November (20). No free API allowance found; billing needs a card, per the platform reading (0). Browser sign-up (0). 0 + 20 + 0 + 0 = 20.

Muse Spark Open 50

Self-hosted rule. Nothing to buy from Meta, and the licence grants use free of charge. The Model as a Service clause asks for a separate agreement only from licensee groups above $20M gross revenue in 12 months (20). Free to run with no card (20). Meta's own weights need a manually reviewed Hugging Face form with legal name and date of birth, or, since the 4 November re-read of Get the model, a signed-in download from Meta's own page, so an agent can't get them alone. The ungated GGUF quantisations from ggml-org and Unsloth that Meta's docs point to download without approval (the Unsloth repository is ungated on 4 November), so part of the line (10 of 20). No payment protocol (0). Hosted copies at $1.25 input and $4.25 output per million tokens need each provider's own account and are not graded. 20 + 20 + 10 + 0 = 50.

Security & auth Muse Ember +20

Muse Ember 64

Platform readings are the meta-model-api listing's of 4 November. Team keys, named, revocable, no scopes (20 of 30). Standard mode only, so content isn't used to train Meta models under the terms, with safety use reserved (17 of 20). Retention 'as needed' with zero retention through sales (10 of 15). Usage dashboard by model, no per-call log (5 of 15). No security.txt on dev.meta.ai, a bug bounty and a security commitments page, no named certification (12 of 20). 20 + 17 + 10 + 5 + 12 = 64.

Muse Spark Open 44

Read as software the owner runs. No account or key to run it. Meta's weights need manual approval on Hugging Face and an access token, and server authentication is the owner's to add (12 of 30). An agent model that calls whatever tools the operator gives it. The card recommends human-in-the-loop confirmation for irreversible actions and Meta reports agentic-risk evaluations, but nothing in the model or template enforces an approval (8 of 20). The chat template neutralises special-token sequences and media placeholders in caller text and wraps tool results in <tool_output>, and the card advises protecting internal assistant messages and layering defences against adversarial input. No indirect prompt-injection results are published (10 of 15). No call log of its own. Visibility depends on the serving stack (5 of 15). Weights are safetensors that load through Transformers' own classes from 5.19, and Meta published cyber, chemical, biological and misalignment evaluations. dev.meta.ai/.well-known/security.txt answers 404 (re-read on 4 November), research.meta.ai per the 15 October check, and no security route is on the card or in the docs, only museoss@meta.com and the Hugging Face discussions (9 of 20). 12 + 8 + 10 + 5 + 9 = 44.

Maintenance & community Muse Spark Open +14

Muse Ember 46

Model reading. Released 9 November 2026 (30). No deprecation policy or notice period found, per the platform reading (0 of 12). Dated launches on the API, the latest Muse Wonder on 12 October and Muse Ember on 9 November (8 of 8). No Model API changelog; a help centre with tickets (8 of 15). No SDK of Meta's own (0 of 15, 0 of 10). 30 + 0 + 8 + 8 + 0 = 46.

Muse Spark Open 60

Read for an open-weight model released three weeks ago. Released on 15 October 2026 (30). One release, BF16 and NVFP4 on the same day, with no history of revisions, successors or notice periods, and Meta Model API has no deprecation policy either (6 of 20). None of the 29 replies counted by script on 4 November across the BF16 repository's discussions came from Meta. Feedback goes to the discussions or museoss@meta.com, and Meta's docs and developer blog carry dated posts (8 of 25). Day-one support in vLLM, SGLang, TensorRT-LLM, Transformers, llama.cpp, MLX and Unsloth with Meta's own recipes, though SGLang's is a dev image and TensorRT-LLM's a feature branch (12 of 15). Current dependencies (Transformers 5.19), but no CI visible to us (4 of 10). 30 + 6 + 8 + 12 + 4 = 60.

Score by category

CategoryWeight this runMuse EmberMuse Spark OpenEdge
Reliability16%204042Muse Spark Open +2
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.26061Muse Spark Open +1
Agent ergonomics13%16.27269Muse Ember +3
Security & auth14%17.56444Muse Ember +20
Payments & pricing10%12.52050Muse Spark Open +30
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.84660Muse Spark Open +14
Transparency & trust7%8.85353even
Negative events≤1500
TotalD 51.8/100D 53.4/100

Facts side by side

FactMuse EmberMuse Spark Open
KindModel APIModel API
VendorMetaMeta
Hosted endpointhttps://api.meta.ai/v1no (local only)
TransportsHTTP
AuthAPI keyNone
PricingPay per useFree
x402nono
LicenceClosed model, used through Meta Model API and Muse Code under the Meta Model API Terms of Service (last updated 9 October 2026). No open weights; Meta's open-weight models are Muse Spark Open and Muse GlimmerMuse Model Licence (source-available weights, not an OSI licence), free of charge with an attribution notice, a separate agreement with Meta before Model as a Service by a licensee group above $20M gross revenue, and the Muse Spark Open Usage Policy
Read-only variant documentednono
llms.txtyesyes
Last release2026-11-092026-10-15
Terms last updated2026-10-022026-10-15
Privacy policy last updatedcouldn't be readno document linked
Customer content may train modelsyesnot found in the text
Terms restrict automated accessnot found in the textnot found in the text
Terms restrict benchmarkingyesnot found in the text
Terms or service can change without noticenot found in the textnot found in the text
Arbitration or class-action waiveryesnot found in the text

How the scores have moved

Verdicts

Muse Ember

A cheap, capable model for the repeated parts of agent work, at $0.10 and $0.50 per million tokens with a 1M context and Meta's figures putting it level with Claude Haiku 5.5 on fewer tokens. Launched today on an API with no readable status page, SLA or deprecation policy, and closed, with no weights.

Muse Spark Open

Muse Spark Open runs on the owner's hardware in seven runtimes, with OpenAI-compatible servers in vLLM and SGLang, standard tool calls, a 1M-token context and Meta's own serving docs. Serving the 537 GB NVFP4 checkpoint takes four B300 GPUs, Meta's download is gated by manual review, and the licence adds attribution and a revenue threshold for model resellers.

Before you call either

Muse Ember

  1. Call https://api.meta.ai/v1/responses (or /chat/completions or /messages) with model muse-ember-1.0 and a bearer MODEL_API_KEY
  2. Set the effort explicitly, reasoning.effort on Responses or reasoning_effort on Chat Completions. It defaults to max, and none isn't accepted; minimal or low suits classification and routing
  3. Keep input under 200K tokens a request, cached tokens and previous_response_id history included, or the whole request bills at the long-context rate
  4. Use it for sub-agents and bounded repeated tasks, and a larger Muse model for the plan, as Meta advises
  5. Use strict structured output for extraction and routing steps
  6. In Claude Code, map the haiku alias and sub-agent model to muse-ember-1.0, per Meta's guide

Muse Spark Open

  1. Start vLLM or SGLang with --reasoning-parser muse --tool-call-parser muse (vLLM also --enable-auto-tool-choice), or tool calls come back as raw text
  2. Set reasoning_effort to low or medium for short tasks. The default is xhigh, reasoning can't be turned off, and other values return an error
  3. Stop only on <|eot|> and end of text. Stopping on <|eom|> cuts off tool calls and final answers
  4. Stream long generations and leave max_tokens room for reasoning, or content can come back empty with finish_reason length
  5. Don't call Meta Model API for this model. Use a self-hosted server or a hosted provider's model string, such as meta-models/muse-spark-open

Questions

Which is better for AI agents, Muse Ember or Muse Spark Open?

Muse Spark Open scores 53.4 (D) on agent readiness against Muse Ember's 51.8 (D), and leads in 4 of 7 scored categories. Muse Ember leads on security & auth.

Do Muse Ember and Muse Spark Open need an API key?

Muse Ember needs an API key. Muse Spark Open needs no key.

Can an agent call Muse Ember and Muse Spark Open without installing anything?

Muse Ember has a hosted endpoint at https://api.meta.ai/v1. No hosted endpoint is listed for Muse Spark Open.

Other comparisons with Muse Ember or Muse Spark Open

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.