BenchLeader

Beam 1.0 501B A23B vs Muse Spark Open

Full profiles: Beam 1.0 501B A23B (Reflection) and Muse Spark Open (Meta): every benchmark, price and rank.

Updated · Data from 10 sources

As of 16 Nov 2026, Beam 1.0 501B A23B and Muse Spark Open are neck and neck on quality (54.2 vs 53.5 on the BenchLeader Index, inside the ±10.0 margin), and Beam 1.0 501B A23B is 3.3× cheaper at $0.600 vs $2.00 per 1M tokens blended. Beam 1.0 501B A23B leads in every category we measure both in.

Key facts (as of 16 Nov 2026)

  • Beam 1.0 501B A23B scores 54.2 on the BenchLeader Index and Muse Spark Open scores 53.5 (BenchLeader, 16 Nov 2026).
  • Beam 1.0 501B A23B costs $0.600 per 1M tokens blended (list price, 16 Nov 2026).
  • Muse Spark Open costs $2.00 per 1M tokens blended (list price, 16 Nov 2026).

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingBeam 1.0 501B A23B52 vs 36 category scoreHigh
ReasoningBeam 1.0 501B A23B62 vs 51 category scoreHigh
Chat quality (human preference)–Only Beam 1.0 501B A23B has an Arena rating (61 category score)Not comparable
Lowest costBeam 1.0 501B A23B$0.475 vs $2.00 per 1M blended; agentic coding session $0.022 vs $0.092–
Fastest first answerBeam 1.0 501B A23B0.3 s (Beam 1.0 501B A23B) vs 9.9 s (Muse Spark Open (xhigh)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Beam 1.0 501B A23B and Muse Spark Open (xhigh) are neck and neck on quality (54.2 vs 53.5, within the ±10.0 margin).
  • Beam 1.0 501B A23B is stronger in coding, reasoning.
  • Beam 1.0 501B A23B is 3.3× cheaper ($0.600 vs $2.00 per 1M blended).

At a glance

SpecBeam 1.0 501B A23BMuse Spark Open
DeveloperReflectionMeta
Released5 Oct 202615 Oct 2026
Context window (tokens)1,048,5761,048,576
Compared atBeam 1.0 501B A23BMuse Spark Open (xhigh)
Open weightsYesYes
Price in / out / cached input per 1M$0.300 / $1.00 / $0.040$1.25 / $4.25 / $0.150

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryBeam 1.0 501B A23BMuse Spark Open (xhigh)Leader
Reasoning62.151.3Beam 1.0 501B A23B
Coding52.036.0Beam 1.0 501B A23B
Agents & tools40.8–not comparable
Maths65.2–not comparable
Knowledge–58.8not comparable
Human preference61.4–not comparable
Multimodal–63.3not comparable
Long context–63.8not comparable

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingAgents & toolsMathsKnowledgeLong contextMultimodalHumanpreferenceBeam 1.0 501B A23B, Reasoning: 62.1Beam 1.0 501B A23B, Coding: 52.0Beam 1.0 501B A23B, Agents & tools: 40.8Beam 1.0 501B A23B, Maths: 65.2Beam 1.0 501B A23B, Human preference: 61.4Muse Spark Open, Reasoning: 51.3Muse Spark Open, Coding: 36.0Muse Spark Open, Knowledge: 58.8Muse Spark Open, Long context: 63.8Muse Spark Open, Multimodal: 63.3

Dashed: that model has no result on the axis it crosses; the line only joins its neighbours.

Not drawn, no result for these models: Instruction following.

Model, click to add or removeIndex

Where sources disagree

Across the 0 benchmarks both have, Beam 1.0 501B A23B leads 0, Muse Spark Open leads 0.

Overall indexBeam 1.0 501B A23BMuse Spark Open (xhigh)Leader
BenchLeader Index54.253.5neck and neck

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureBeam 1.0 501B A23BMuse Spark Open (xhigh)
Blended price per 1M tokens$0.475$2.00
Input / output per 1M$0.300 / $1.00$1.25 / $4.25
Cached input per 1M$0.040$0.150
Price per task: chat reply$0.0004$0.0018
Price per task: summarize a 30-page report$0.0042$0.018
Price per task: code edit$0.0033$0.014
Price per task: agentic coding session$0.022$0.092
Price per task: structured extraction$0.0008$0.0034
Agentic coding session, input mostly cached$0.010$0.043
Cost per benchmark task (AA Index)–$2.02
Output tokens per task (AA Index)–66k
Of which reasoning–32k
Time per task (AA Index)–6 min
Output speed193 tok/s180 tok/s
Time to first answer0.3 s9.9 s
Time to a complete answer, last measured 13 Nov 2026–39.9 s

Each model is compared at its best-ranked reasoning-effort setting. Beam 1.0 501B A23B scores 54.2 and takes 0 s to its first answer token. Muse Spark Open (xhigh) scores 53.5 and takes 10 s to its first answer token. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

85 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherBeam 1.0 501B A23BMuse Spark Open (xhigh)
BenchLeader IndexBenchLeader54.253.5
Agents & tools scoreBenchLeader40.8–
Coding scoreBenchLeader52.036.0
Human preference scoreBenchLeader61.4–
Maths scoreBenchLeader65.2–
Reasoning scoreBenchLeader62.151.3
Knowledge scoreBenchLeader–58.8
Long context scoreBenchLeader–63.8
Multimodal scoreBenchLeader–63.3
Blended price $/MBenchLeader$0.475$2.00
Output speedBenchLeader193 tok/s180 tok/s
Time to first answerBenchLeader0.3 s9.9 s
Context windowBenchLeader1,048,576 tokens1,048,576 tokens
Reasoning
Arena Hard PromptsArena (LMArena)1461–
LiveBench 2.0 ReasoningLiveBench–73.7%†
GPQA Diamond (AA)Artificial Analysis–89.0%
Humanity's Last Exam (AA)Artificial Analysis–43.9%
HLE-Diamond (AA)Artificial Analysis–26.3%
ARC-AGI-1ARC Prize–89.0%
ARC-AGI-2ARC Prize–57.5%
ARC-AGI-3ARC Prize–0.4%
CritPtArtificial Analysis–14.9%
CritPt v2 hard subset (AA)Artificial Analysis–29.5%
HLE-Diamond (Scale)Scale AI SEAL–28.0%†
Arena ExpertArena (LMArena)1481–
Coding
Arena CodingArena (LMArena)1505–
Arena WebDevArena (LMArena)14371526†
LiveBench 2.0 CodingLiveBench–55.1%†
SciCode (AA)Artificial Analysis–58.5%
Code MigrationVals AI–21.2%
Code Migration v1.1Vals AI–26.9%
Vibe Code Bench v1.1Vals AI–76.2%
SWE Atlas v1.1Scale AI SEAL–42.3%†
SWE Atlas v1.1: Codebase QnAScale AI SEAL–50.0%†
SWE Atlas v1.1: Test WritingScale AI SEAL–41.1%†
SWE Atlas v1.1: RefactoringScale AI SEAL–35.7%†
SWE-Bench Pro V2Scale AI SEAL–95.0%
SWE-Bench Pro V2 HARDScale AI SEAL–82.3%
Show all 85 benchmarks
MetricPublisherBeam 1.0 501B A23BMuse Spark Open (xhigh)
Coding
Terminal-Bench 4.0 (AA)Artificial Analysis–8.1%
Terminal-Bench 2.1 (AA)Artificial Analysis–76.4%
GameBench 2 OverallDuelLab GameBench 242.855.3†
GameBench 2 by reasoning settingDuelLab GameBench 239.1†59.3
Design Arena WebsiteDesign Arena–1305
FrontierCode 1.1 (AA)Artificial Analysis–27.6%
FrontierCode 2.0 (AA)Artificial Analysis–7.6%
Agents & tools
Arena AgentArena (LMArena)-5.9–
LiveBench Long AgenticLiveBench–12.1%†
LiveBench BrowserLiveBench–26.2%†
HiL-BenchScale AI SEAL–59.7%†
GDPval-AA v2.1, normalised scoreArtificial Analysis–45.1%
GDPval-AA v2.1Artificial Analysis–1401
Finance Agent v2Vals AI–61.3%
Finance Agent v2.1Vals AI–60.7%
Harvey's Legal Agent Benchmark v1.1Vals AI–25.0%
Harvey's Legal Agent Benchmark v1Vals AI–16.7%
Legal Research BenchVals AI–38.5%
Tax Agent BenchVals AI–26.3%
Terminal-Bench 4.0 (Vals)Vals AI–6.6%
Arena Agent Task CompletionArena (LMArena)-9.2–
AutomationBenchArtificial Analysis–49.7%
GDP.pdf (AA)Artificial Analysis–18.6%
AA-Briefcase v1.1Artificial Analysis–1280
GDPxGDPx by SemiAnalysis–1.7%†
Terminal-Bench-Science 0.1 (AA)Artificial Analysis–1.9%
Maths
ArXivMathMathArena–27.0%
BrokenArXivMathArena–30.1%
Arena MathsArena (LMArena)1471–
Knowledge
LiveBench 2.0 Data AnalysisLiveBench–60.2%†
LiveBench ResearchLiveBench–82.7%†
AA-Omniscience v1.1Artificial Analysis–25.8
PRBench LegalScale AI SEAL–59.8%†
Excel Modeling BenchmarkVals AI–53.7%
Excel Modeling Benchmark v1.1Vals AI–62.0%
AA-Omniscience v1.1: accuracyArtificial Analysis–42.0%
AA-Omniscience v1.1: non-hallucinationArtificial Analysis–72.0%
Instruction following
MultiChallengeScale AI SEAL–74.2%†
Arena Instruction FollowingArena (LMArena)1433–
Human preference
Arena TextArena (LMArena)1434–
Arena Creative WritingArena (LMArena)1388–
Arena Multi-turnArena (LMArena)1432–
Arena Longer QueriesArena (LMArena)1438–
Multimodal
LiveBench DesignLiveBench–58.1%†
MMMU-ProArtificial Analysis–80.9%
SAGEVals AI–85.8%
Humanity's Sixth SenseScale AI SEAL–29.6%†
Long context
AA-LCR v1.1Artificial Analysis–78.3%
MLCRArtificial Analysis–7.8%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub–155.4†
LiveBench 2.0LiveBench–48.5%†
AA Intelligence Index v5.0.1Artificial Analysis–24.9
AA Intelligence Index v4.3.2Artificial Analysis–38.9
AA Progress Index v1Artificial Analysis–306.7
Vals IndexVals AI–48.6
Vals Finance IndexVals AI–53.0%
Vals Legal IndexVals AI–68.5%
Safety & honesty
FORTRESSScale AI SEAL11.9%7.6%†
DistressBenchScale AI SEAL–87.7%†
Metadata
AA Openness IndexArtificial Analysis–27.8

Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • Beam 1.0 501B A23B: Beam 1.0 501B A23B input price moved from $0.40/M to $0.30/M (25% down); Beam 1.0 501B A23B output price moved from $1.20/M to $1.00/M (17% down).
  • Beam 1.0 501B A23B: 1 score revision.
  • Muse Spark Open (xhigh): 1 score revision.
  • Beam 1.0 501B A23B: BenchLeader Index 54.9 → 54.2.
  • Beam 1.0 501B A23B: 2 score revisions.
  • Beam 1.0 501B A23B: 1 new result (FORTRESS); 1 score revision.
  • Beam 1.0 501B A23B: 1 score revision.
  • Muse Spark Open (xhigh): 1 new result (GameBench 2 by reasoning setting); 1 score revision.
  • Muse Spark Open (xhigh): BenchLeader Index 53.3 → 53.8.
  • Beam 1.0 501B A23B: 1 score revision.

Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs

Beam 1.0 501B A23B vs Muse Spark Open: questions

Is Beam 1.0 501B A23B better than Muse Spark Open?
Beam 1.0 501B A23B and Muse Spark Open (xhigh) are neck and neck on quality (54.2 vs 53.5, within the ±10.0 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 16 Nov 2026, so check the category scores for your use.
Beam 1.0 501B A23B vs Muse Spark Open: which is cheaper, and will that change?
Beam 1.0 501B A23B is cheaper today: $0.600 against $2.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Beam 1.0 501B A23B or Muse Spark Open better for coding?
Beam 1.0 501B A23B scores higher in coding (52 vs 36 on the category index, where 50 is average).
Which is faster, Beam 1.0 501B A23B or Muse Spark Open?
Beam 1.0 501B A23B streams faster (193 vs 180 output tokens per second, Artificial Analysis 7-day median).
Which is better for long documents?
Both accept 1,048,576 tokens of context.
More questions (2)
Which has lower latency?
Beam 1.0 501B A23B answers first: 0.3 s against 9.9 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
When was Muse Spark Open released?
Muse Spark Open was released on 15 Oct 2026.