Muse Ember vs Muse Spark Open
Muse Ember leads Muse Spark Open on coding, 64.9 to 43.4 on the Vibe Coding Index, and Muse Ember costs 8.5x less per output token ($0.50 vs $4.25 per million). Updated November 14, 2026; 15 benchmarks have a result for both.
Muse Ember or Muse Spark Open: which to pick
Pick Muse Ember for coding: it scores 64.9 on the Vibe Coding Index against 43.4 for Muse Spark Open, though on 3 benchmarks against 4.
Pick Muse Spark Open for work measured by HLE-Diamond (Artificial Analysis run) or AA-Omniscience, where it leads.
- Shared results
- 15
- Muse Ember only
- 0
- Muse Spark Open only
- 32
- Like-for-like categories
- 5 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
Muse Ember
Muse Ember scores 64.9 against 43.4.
3 and 4 benchmarks behind the scores
Intelligence
AA Intelligence Index
No call
Only Muse Spark Open has a AA Intelligence Index result, 24.9.
Directional only
Web design
Design Arena: websites
No call
Only Muse Spark Open has a Design Arena: websites result, 1306.
Directional only
General
Vals Index
Muse Ember
Muse Ember leads, 60.5% against 48.6%.
Same board, both measured
Long documents
Listed context window
No call
Both take 1.05M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Muse Ember
Muse Ember $0.0003, Muse Spark Open $0.0034 per turn.
List prices
At a glance
- Coding index
- Muse Ember 64.9 from 3 benchmarks, Muse Spark Open 43.4 from 4 benchmarks.
- Widest category
- Coding, 21.3 points apart on average over 7 shared benchmarks.
- Widest single gap
- Terminal-Bench 4 (Vals AI): Muse Ember 39.4% against Muse Spark Open 6.6%, 43.8 points apart on a 0 to 100 scale.
- Output price
- Muse Ember $0.50 against Muse Spark Open $4.25 per million output tokens, 8.5x apart.
- Context window
- Both take 1.05M tokens.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Muse Ember | Bars | Muse Spark Open | Ties |
|---|---|---|---|---|
| Coding | 7 | 0 | ||
| Agentic work | 3 | 0 | 1 tie | |
| Reasoning | 1 | 1 | ||
| Knowledge | 0 | 1 | ||
| General | 1 | 0 | ||
| All shared boards | 12 | 2 | 1 tie |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index70
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| LiveBench coding | Not measured | 55.1% | Only Muse Spark Open measured |
| LMArena WebDev | 1609 | 1526 | Muse Ember |
| Terminal-Bench 4 (Vals AI) | 39.4% | 6.6% | Muse Ember |
| Vibe Code Bench (Vals AI) | 83.7% | 76.2% | Muse Ember |
| Code migration (Vals AI) | Not measured | 21.2% | Only Muse Spark Open measured |
| SciCode | Not measured | 58.4% | Only Muse Spark Open measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 40.9% | 8.1% | Muse Ember |
| FrontierCode 2.0 | 27.2% | 7.6% | Muse Ember |
| ALE-Bench | Not measured | 1590.1 | Only Muse Spark Open measured |
| LMArena WebDev: React | 1616 | 1523 | Muse Ember |
| LMArena WebDev: HTML | 1621 | 1475 | Muse Ember |
| SWE-Bench Pro V2 | Not measured | 95.0% | Only Muse Spark Open measured |
| GameBench 2 | Not measured | 59.3 | Only Muse Spark Open measured |
| Neither model has a result on the other 39 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 24.9 | Only Muse Spark Open measured |
| Epoch Capabilities Index | Not measured | 155.4 | Only Muse Spark Open measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| LiveBench research | Not measured | 82.7% | Only Muse Spark Open measured |
| LiveBench browser | Not measured | 26.2% | Only Muse Spark Open measured |
| LiveBench long agentic | Not measured | 12.1% | Only Muse Spark Open measured |
| Finance Agent (Vals AI) | Not measured | 61.3% | Only Muse Spark Open measured |
| Tax Agent Bench (Vals AI) | 38.2% | 26.3% | Muse Ember |
| Legal Research Bench (Vals AI) | 49.0% | 38.5% | Muse Ember |
| Harvey's Legal Agent Benchmark (Vals AI) | 25.0% | 25.0% | Tie |
| Excel Modeling Benchmark (Vals AI) | Not measured | 53.7% | Only Muse Spark Open measured |
| APEX-Agents | 64.3% | 33.4% | Muse Ember |
| Neither model has a result on the other 35 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| LiveBench reasoning | Not measured | 73.7% | Only Muse Spark Open measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 89.0% | Only Muse Spark Open measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 43.9% | Only Muse Spark Open measured |
| HLE-Diamond (Artificial Analysis run) | 25.1% | 26.3% | Muse Spark Open |
| CritPt v2, hard subset (Artificial Analysis run) | 39.0% | 29.5% | Muse Ember |
| ARC-AGI | Not measured | 89.0% | Only Muse Spark Open measured |
| ARC-AGI-2 | Not measured | 57.5% | Only Muse Spark Open measured |
| Humanity's Last Exam (Scale run) | Not measured | 28.0% | Only Muse Spark Open measured |
| Neither model has a result on the other 18 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| MathArena | Not measured | 25.6% | Only Muse Spark Open measured |
| Neither model has a result on the other 13 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| AA-Omniscience | 7.7 | 25.8 | Muse Spark Open |
| PRBench Legal | Not measured | 59.8% | Only Muse Spark Open measured |
| Neither model has a result on the other 19 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1306 | Only Muse Spark Open measured |
| Design Arena: UI components | Not measured | 1312 | Only Muse Spark Open measured |
| Design Arena: game dev | Not measured | 1265 | Only Muse Spark Open measured |
| Design Arena: 3D | Not measured | 1247 | Only Muse Spark Open measured |
| Design Arena: all code categories | Not measured | 1301 | Only Muse Spark Open measured |
| LiveBench design | Not measured | 58.1% | Only Muse Spark Open measured |
| Neither model has a result on the other 3 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| Vals Index | 60.5% | 48.6% | Muse Ember |
| LiveBench data analysis | Not measured | 60.2% | Only Muse Spark Open measured |
| SAGE (Vals AI) | Not measured | 85.8% | Only Muse Spark Open measured |
| Long context reasoning (AA-LCR) | Not measured | 78.3% | Only Muse Spark Open measured |
| GDPval-AA | Not measured | 1401 | Only Muse Spark Open measured |
| MultiChallenge | Not measured | 74.2% | Only Muse Spark Open measured |
| Neither model has a result on the other 43 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Muse Ember | Muse Spark Open | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Muse Ember costs less
Repository review
50K input + 3K output tokens
Muse Ember costs less
Agent loop
200K input + 10K output tokens
Muse Ember costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Meta Muse Ember | Meta Muse Spark Open |
|---|---|---|
| Maker | Meta | Meta |
| Weights | No public weights listed | Open weights |
| Released | November 9, 2026 | October 15, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $0.10 | $1.25 |
| Output $/Mtok | $0.50 | $4.25 |
| Coding index | 64.9 | 43.4 |
| Independent measurements behind it | 3 | 4 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Muse Ember better than Muse Spark Open for coding?
Pick Muse Ember for coding: it scores 64.9 on the Vibe Coding Index against 43.4 for Muse Spark Open, though on 3 benchmarks against 4. Pick Muse Spark Open for work measured by HLE-Diamond (Artificial Analysis run) or AA-Omniscience, where it leads.
Which is cheaper, Muse Ember or Muse Spark Open?
Muse Ember costs $0.10 per million input tokens and $0.50 per million output tokens; Muse Spark Open costs $1.25 and $4.25. Cost per solved task is known for neither of them.
Which has the bigger context window, Muse Ember or Muse Spark Open?
Neither: both take 1.05M tokens.
How many benchmarks is the Muse Ember vs Muse Spark Open comparison based on?
15 benchmarks have a result for both models as of November 14, 2026. Muse Ember's coding index rests on 3 independent benchmarks and Muse Spark Open's on 4. Where only one model has a result, the page says so instead of naming a winner.