Head-to-head comparison

Muse Spark 1.3 vs. GLM-5.3-Flash

Compare published benchmark results from matching versions and cohorts, API pricing, and technical specifications. Any documented test setup differences remain visible.

shared result series: 202
distinct benchmarks: 18
latest retrieval: 2026-10-15

First modelSecond model

Specifications

General specifications

Release date, availability, context window, architecture, and published parameter counts at a glance. Undisclosed details remain marked as unavailable.

Specifications for Muse Spark 1.3 and GLM-5.3-Flash
Attribute	Muse Spark 1.3	GLM-5.3-Flash
Provider	Meta	Z.ai
Released	Sep 2, 2026	Aug 26, 2026
Availability	Active	Open Source
Model type	Proprietary	Open Weights
Parameters	Not published	320B, 18B active
Architecture	Not published	Mixture of Experts
Context window	1,048,576 tokens	1,000,000 tokens
Knowledge cutoff	Not published	Not published
Technical sources	Model, Context	Model, Context

Pricing

API pricing compared

The table compares official standard API rates per 1M tokens, including published cache pricing. Batch, fast, priority, regional surcharges, tool calls, and cloud platform pricing are excluded.

API pricing for Muse Spark 1.3 and GLM-5.3-Flash
Price type	Muse Spark 1.3	GLM-5.3-Flash
API input	$1.25	$0.15
API output	$4.25	$0.5
Cache read	$0.15	$0.03
Cache write (5 min.)	Not listed	Not listed
Cache write (1 hr.)	Not listed	Not listed
Price verified	10/18/2026 Source	09/22/2026 Source

Muse Spark 1.3GLM-5.3-Flash

API inputUSD per 1M tokens, base tier

Muse Spark 1.3$1.25

GLM-5.3-Flash$0.15

API outputUSD per 1M tokens, base tier

Muse Spark 1.3$4.25

GLM-5.3-Flash$0.5

Cache readUSD per 1M tokens, base tier

Muse Spark 1.3$0.15

GLM-5.3-Flash$0.03

Key to the marks:better valuebetter, but not conclusivetie

Pricing by provider

Direct developer pricing and provider endpoint tariffs routed through OpenRouter are listed separately. Prices are in USD per 1M tokens. Context-dependent direct tiers plus regional, Flex, Priority, and other OpenRouter tariffs remain individually identifiable.

Provider pricing for Muse Spark 1.3 and GLM-5.3-Flash
Model	Provider	Purchase route and tariff	Input per 1M tokens	Output per 1M tokens	Source
Muse Spark 1.3	Meta	Via OpenRoutermeta	Lowest listed price: $1.25	Lowest listed price: $4.25	Source10/18/2026
GLM-5.3-Flash	Z.ai	Direct from developerStandard API	$0.15	$0.5	Source09/22/2026
	AtlasCloud	Via OpenRouteratlas-cloud/fp8	$0.15	$0.5	Source10/18/2026
	BaseTen	Via OpenRouterbaseten/fp8	$0.15	$0.5	Source10/18/2026
	BaseTen	Via OpenRouterbaseten/fp8/us	$0.15	$0.5	Source10/18/2026
	CoreWeave	Via OpenRoutercoreweave/nvfp4	$0.15	$0.5	Source10/18/2026
	Crusoe	Via OpenRoutercrusoe/fp4	$0.15	$0.5	Source10/18/2026
	Decart	Via OpenRouterdecart/fp4	$0.0825	$0.275	Source10/18/2026
	DeepInfra	Via OpenRouterdeepinfra/fp4	$0.075	$0.25	Source10/18/2026
	DigitalOcean	Via OpenRouterdigitalocean	$0.1312	$0.4375	Source10/18/2026
	Fireworks	Via OpenRouterfireworks	$0.15	$0.5	Source10/18/2026
	Fireworks	Via OpenRouterfireworks/us	$0.225	$0.75	Source10/18/2026
	Friendli	Via OpenRouterfriendli	$0.15	$0.5	Source10/18/2026
	GMICloud	Via OpenRoutergmicloud/fp8	$0.09	$0.3	Source10/18/2026
	Inceptron	Via OpenRouterinceptron/fp8	$0.1	$0.69	Source10/18/2026
	InferenceNet	Via OpenRouterinference-net/fp4	$0.12	$0.388	Source10/18/2026
	Makora	Via OpenRoutermakora	$0.14	$0.47	Source10/18/2026
	Modal	Via OpenRoutermodal/nvfp4	$0.15	$0.5	Source10/18/2026
	Morph	Via OpenRoutermorph/fp8	$0.129	$0.65	Source10/18/2026
	Near AI	Via OpenRouternear-ai/fp8	$0.08	$0.5	Source10/18/2026
	NextBit	Via OpenRouternextbit/fp4	$0.25	$0.7	Source10/18/2026
	Novita	Via OpenRouternovita/fp8	$0.084	$0.28	Source10/18/2026
	OpenInference	Via OpenRouteropen-inference/fp4	Lowest listed price: $0.048	Lowest listed price: $0.2	Source10/18/2026
	Parasail	Via OpenRouterparasail/fast	$0.1875	$0.625	Source10/18/2026
	Parasail	Via OpenRouterparasail/fp4	$0.15	$0.5	Source10/18/2026
	Phala	Via OpenRouterphala/nvfp4	$0.1275	$0.425	Source10/18/2026
	PrimeIntellect	Via OpenRouterprimeintellect	$0.15	$0.5	Source10/18/2026
	Reka	Via OpenRouterreka	$0.06	$1.6	Source10/18/2026
	Relace	Via OpenRouterrelace	$0.075	$0.5	Source10/18/2026
	Sail Research	Via OpenRoutersail-research/fp4	$0.0784	$0.3	Source10/18/2026
	Sail Research	Via OpenRoutersail-research/us	$0.0784	$0.3	Source10/18/2026
	SiliconFlow	Via OpenRoutersiliconflow/fp8	$0.15	$0.5	Source10/18/2026
	StreamLake	Via OpenRouterstreamlake/fp8	$0.075	$0.25	Source10/18/2026
	Together	Via OpenRoutertogether	$0.15	$0.5	Source10/18/2026
	Venice	Via OpenRoutervenice	$0.15	$0.5	Source10/18/2026
	Wafer	Via OpenRouterwafer/nvfp4	$0.12	$0.5	Source10/18/2026
	Z.AI	Via OpenRouterz-ai/fp8	$0.15	$0.5	Source10/18/2026

Performance

Shared benchmarks

Only published results with the same benchmark version, task, metric, and comparison cohort are paired. Different reasoning levels, agents, harnesses, or output limits appear directly in the table.

Capability profile from matched benchmarks

Each axis averages directly matched benchmark families. A value of 100 means the stronger value within this pair, not a universal quality score.

Not enough directly matched categories for a reliable radar

Only 0 of at least 5 categories have comparable, direct model measurements. The benchmark tables below show the available evidence without filling missing axes.

Muse Spark 1.3GLM-5.3-Flash

Code MigrationOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.327.58 %

GLM-5.3-Flash20.515 %

CyberBench v1.1Overall. Higher is better. Test setup not documented as identical

Muse Spark 1.369.405 %

GLM-5.3-Flash67.857 %

EMBOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.362.712 %

GLM-5.3-Flash55.932 %

Finance Agent (v2)Overall. Higher is better. Test setup not documented as identical

Muse Spark 1.358.901 %

GLM-5.3-Flash57.85 %

Harvey's Legal Agent BenchmarkOverall · Task fully resolved. Higher is better. Test setup not documented as identical

Muse Spark 1.322.917 %

GLM-5.3-Flash6.667 %

IOIOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.343.944 %

GLM-5.3-Flash52.5 %

Legal Research BenchOverall · All-pass. Higher is better. Test setup not documented as identical

Muse Spark 1.340.865 %

GLM-5.3-Flash45.192 %

LMArena Code Arena WebDevOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.31,621.927

GLM-5.3-Flash1,606.051

LMArena Text, Style ControlOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.31,494.213

GLM-5.3-Flash1,474.228

LMArena Vision, Style ControlOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.31,286.221

GLM-5.3-Flash1,273.331

ProgramBenchOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.30.5 %

GLM-5.3-Flash0 %

ProofBenchOverall. Higher is better. Test setup not documented as identical

Muse Spark 1.355 %

GLM-5.3-Flash21 %

Key to the marks:better valuebetter, but not conclusivetie

Benchmark scores for Muse Spark 1.3 and GLM-5.3-Flash
Benchmark and task	Muse Spark 1.3	GLM-5.3-Flash	Source
Code MigrationOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value27.58 %	20.515 %	Vals AI Source details Data as of: 10/12/2026
CyberBench v1.1Overall Test details Test setup not documented as identicalNo direct winnerVersion: 1.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value69.405 %	67.857 %	Vals AI Source details Data as of: 10/12/2026
EMBOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value62.712 %	55.932 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)Overall Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value58.901 %	57.85 %	Vals AI Source details Data as of: 10/12/2026
Harvey's Legal Agent BenchmarkOverall · Task fully resolved Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: task resolution rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value22.917 %	6.667 %	Vals AI Source details Data as of: 10/12/2026
IOIOverall Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	43.944 %	Higher measured value52.5 %	Vals AI Source details Data as of: 10/14/2026
Legal Research BenchOverall · All-pass Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: all-pass rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	40.865 %	Higher measured value45.192 %	Vals AI Source details Data as of: 10/12/2026
LMArena Code Arena WebDevOverall Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3 (xHigh), Reasoning: xhigh, Harness: Arena Code Arena WebDev GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Code Arena WebDevMuse Spark 1.3: Arena rating from 7,987 published pairwise votes.GLM-5.3-Flash: Arena rating from 16,470 published pairwise votes.	Higher measured value1,621.927 95% confidence interval: ±8.096The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,606.051 95% confidence interval: ±6.732The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Text, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Text, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Text, style controlMuse Spark 1.3: Arena rating from 16,731 published pairwise votes.GLM-5.3-Flash: Arena rating from 31,379 published pairwise votes.	Higher measured value1,494.213 95% confidence interval: ±5.665The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,474.228 95% confidence interval: ±4.913The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3 (xHigh), Reasoning: xhigh, Harness: Arena Vision, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Vision, style controlMuse Spark 1.3: Arena rating from 4,077 published pairwise votes.GLM-5.3-Flash: Arena rating from 11,157 published pairwise votes.	Higher measured value1,286.221 95% confidence interval: ±10.243The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,273.331 95% confidence interval: ±7.077The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
ProgramBenchOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: fully resolved rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value0.5 %	0 %	Vals AI Source details Data as of: 10/12/2026
ProofBenchOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value55 %	21 %	Vals AI Source details Data as of: 10/12/2026
Tax Agent BenchOverall Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value35.974 %	31.55 %	Vals AI Source details Data as of: 10/12/2026
Terminal-Bench 2.1Overall Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value72.285 %	62.921 %	Vals AI Source details Data as of: 09/28/2026
Terminal-Bench 4.0Overall Test details Test setup not documented as identicalNo direct winnerVersion: 4.0Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	10.606 %	Higher measured value37.374 %	Vals AI Source details Data as of: 10/13/2026
Terminal-Bench ScienceOverall Test details Test setup not documented as identicalNo direct winnerVersion: 0.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Harness: Mini-SWE-agent, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Harness: Mini-SWE-agent, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value4.286 %	2.857 %	Vals AI Source details Data as of: 10/14/2026
Vals IndexOverall Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: weighted index scoreScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value53.197 %	44.581 %	Vals AI Source details Data as of: 10/13/2026
Vibe Code Bench v1.1Overall Test details Test setup not documented as identicalNo direct winnerVersion: 1.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Harness: OpenHands, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Harness: OpenHands, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value82.859 %	30.759 %	Vals AI Source details Data as of: 10/12/2026
Code MigrationCLI Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value13.916 %	7.354 %	Vals AI Source details Data as of: 10/12/2026
CyberBench v1.1Patch Test details Test setup not documented as identicalNo direct winnerVersion: 1.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	82.143 %	Higher measured value85.714 %	Vals AI Source details Data as of: 10/12/2026
EMBComps Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value64.798 %	44.602 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)Adjustments Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	52.818 %	Higher measured value53.193 %	Vals AI Source details Data as of: 10/12/2026
Harvey's Legal Agent BenchmarkAntitrust Competition · Criteria passed Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: criteria pass rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value96.377 %	86.594 %	Vals AI Source details Data as of: 10/12/2026
IOIIOI 2024 Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	42.667 %	Higher measured value53.167 %	Vals AI Source details Data as of: 10/14/2026
Legal Research BenchAdministrative / Regulatory · All-pass Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: all-pass rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	51.515 %	Higher measured value57.576 %	Vals AI Source details Data as of: 10/12/2026
LMArena Code Arena WebDevOverall Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Code Arena WebDev GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Code Arena WebDevMuse Spark 1.3: Arena rating from 8,683 published pairwise votes.GLM-5.3-Flash: Arena rating from 16,470 published pairwise votes.	Higher measured value1,658.54 95% confidence interval: ±8.09The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,606.051 95% confidence interval: ±6.732The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Text, Style ControlBusiness, management and finance Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Text, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Text, style controlMuse Spark 1.3: Arena rating from 3,041 published pairwise votes.GLM-5.3-Flash: Arena rating from 5,788 published pairwise votes.	Higher measured value1,503.61 95% confidence interval: ±11.102The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,470.347 95% confidence interval: ±8.428The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlOverall Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Vision, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Vision, style controlMuse Spark 1.3: Arena rating from 4,550 published pairwise votes.GLM-5.3-Flash: Arena rating from 11,157 published pairwise votes.	Higher measured value1,289.717 95% confidence interval: ±9.808The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,273.331 95% confidence interval: ±7.077The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
ProgramBenchAlmost Resolved Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: almost-resolved rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value13 %	7 %	Vals AI Source details Data as of: 10/12/2026
Tax Agent BenchControversy & Precedence Analysis Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value14.667 %	10 %	Vals AI Source details Data as of: 10/12/2026
Terminal-Bench 2.1Easy Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	75 %	Higher measured value91.667 %	Vals AI Source details Data as of: 09/28/2026
Terminal-Bench 4.0Hardware Test details Test setup not documented as identicalNo direct winnerVersion: 4.0Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	0 %	Higher measured value20 %	Vals AI Source details Data as of: 10/13/2026
Terminal-Bench ScienceEarth Sciences Test details Test setup not documented as identicalNo direct winnerVersion: 0.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Harness: Mini-SWE-agent, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Harness: Mini-SWE-agent, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	No direct winner0 %	No direct winner0 %	Vals AI Source details Data as of: 10/14/2026
Vals IndexCode Migration Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: benchmark scoreScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value42.192 %	20.185 %	Vals AI Source details Data as of: 10/13/2026
Code MigrationCOBOL Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value68.571 %	60 %	Vals AI Source details Data as of: 10/12/2026
CyberBench v1.1PoC Test details Test setup not documented as identicalNo direct winnerVersion: 1.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value56.667 %	50 %	Vals AI Source details Data as of: 10/12/2026
EMBDataroom Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value84.191 %	65.712 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)All-Pass Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value48.737 %	46.259 %	Vals AI Source details Data as of: 10/12/2026
Harvey's Legal Agent BenchmarkAntitrust Competition · Task fully resolved Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: task resolution rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value20 %	0 %	Vals AI Source details Data as of: 10/12/2026
IOIIOI 2025 Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	46.833 %	Higher measured value56.667 %	Vals AI Source details Data as of: 10/14/2026
Legal Research BenchAdministrative / Regulatory · Weighted partial credit Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: weighted partial-credit scoreScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	85.884 %	Higher measured value90.822 %	Vals AI Source details Data as of: 10/12/2026
LMArena Code Arena WebDevImage to webdev Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3 (xHigh), Reasoning: xhigh, Harness: Arena Code Arena WebDev GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Code Arena WebDevMuse Spark 1.3: Arena rating from 2,152 published pairwise votes.GLM-5.3-Flash: Arena rating from 4,315 published pairwise votes.	Higher measured value1,597.944 95% confidence interval: ±13.55The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,580.658 95% confidence interval: ±10.006The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
LMArena Text, Style ControlChinese Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Text, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Text, style controlMuse Spark 1.3: Arena rating from 1,298 published pairwise votes.GLM-5.3-Flash: Arena rating from 2,386 published pairwise votes.	1,529.417 95% confidence interval: ±16.859The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	Higher measured value1,530.199 95% confidence interval: ±12.82The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlChinese Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3 (xHigh), Reasoning: xhigh, Harness: Arena Vision, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Vision, style controlMuse Spark 1.3: Arena rating from 226 published pairwise votes.GLM-5.3-Flash: Arena rating from 453 published pairwise votes.	Higher measured value1,360.609 95% confidence interval: ±38.015The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,342.756 95% confidence interval: ±28.033The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
ProgramBenchFully Resolved Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: fully resolved rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value0.5 %	0 %	Vals AI Source details Data as of: 10/12/2026
Tax Agent BenchControversy & Precedence Analysis Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value60.859 %	52.416 %	Vals AI Source details Data as of: 10/12/2026
Terminal-Bench 2.1Hard Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value61.111 %	38.889 %	Vals AI Source details Data as of: 09/28/2026
Terminal-Bench 4.0Media Test details Test setup not documented as identicalNo direct winnerVersion: 4.0Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	0 %	Higher measured value25 %	Vals AI Source details Data as of: 10/13/2026
Terminal-Bench ScienceEngineering Sciences Test details Test setup not documented as identicalNo direct winnerVersion: 0.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Harness: Mini-SWE-agent, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Harness: Mini-SWE-agent, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	No direct winner0 %	No direct winner0 %	Vals AI Source details Data as of: 10/14/2026
Vals IndexEMB Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: benchmark scoreScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value62.712 %	55.932 %	Vals AI Source details Data as of: 10/13/2026
Code MigrationCode Quality Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: mean reviewer code-quality ratingScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	4.02 / 5	Higher measured value4.355 / 5	Vals AI Source details Data as of: 10/12/2026
EMBDCF Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	55.697 %	Higher measured value59.852 %	Vals AI Source details Data as of: 10/12/2026
Finance Agent (v2)Comparables Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value47.167 %	45.936 %	Vals AI Source details Data as of: 10/12/2026
Harvey's Legal Agent BenchmarkArbitration International Dispute Resolution · Criteria passed Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: criteria pass rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value93.907 %	89.606 %	Vals AI Source details Data as of: 10/12/2026
IOIIOI 2026 Test details Test setup not documented as identicalNo direct winnerVersion: 2Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	42.333 %	Higher measured value47.667 %	Vals AI Source details Data as of: 10/14/2026
Legal Research BenchBusiness & Commercial · All-pass Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: all-pass rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	32.558 %	Higher measured value44.186 %	Vals AI Source details Data as of: 10/12/2026
LMArena Code Arena WebDevImage to webdev Test details Test setup not documented as identicalNo direct winnerVersion: webdevMetric: WebDev Arena Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Code Arena WebDev GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Code Arena WebDevMuse Spark 1.3: Arena rating from 4,688 published pairwise votes.GLM-5.3-Flash: Arena rating from 4,315 published pairwise votes.	Higher measured value1,641.977 95% confidence interval: ±9.755The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,580.658 95% confidence interval: ±10.006The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
LMArena Text, Style ControlCoding Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Text, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Text, style controlMuse Spark 1.3: Arena rating from 4,357 published pairwise votes.GLM-5.3-Flash: Arena rating from 8,116 published pairwise votes.	Higher measured value1,536.661 95% confidence interval: ±9.472The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	1,525.054 95% confidence interval: ±7.559The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/13/2026
LMArena Vision, Style ControlChinese Test details Test setup not documented as identicalNo direct winnerVersion: style-control-v1Metric: style-controlled Bradley-Terry ratingScoring: Higher is betterMuse Spark 1.3: Snapshot: muse-spark-1.3-max, Reasoning: max, Harness: Arena Vision, style control GLM-5.3-Flash: Snapshot: glm-5.3-flash, Harness: Arena Vision, style controlMuse Spark 1.3: Arena rating from 253 published pairwise votes.GLM-5.3-Flash: Arena rating from 453 published pairwise votes.	1,342.141 95% confidence interval: ±37.194The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	Higher measured value1,342.756 95% confidence interval: ±28.033The Arena score measures human preference rather than a fixed capability test. The confidence interval, Arena category, and model configuration are part of the result.	LMArena Source details Data as of: 10/09/2026
ProgramBenchRaw Pass Rate Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: raw test pass rateScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value71.722 %	48.609 %	Vals AI Source details Data as of: 10/12/2026
Tax Agent BenchCurrent & Temporal Analysis Test details Test setup not documented as identicalNo direct winnerVersion: 1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value49.769 %	43.197 %	Vals AI Source details Data as of: 10/12/2026
Terminal-Bench 2.1Medium Test details Test setup not documented as identicalNo direct winnerVersion: 2.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	Higher measured value78.182 %	73.939 %	Vals AI Source details Data as of: 09/28/2026
Terminal-Bench 4.0ML Test details Test setup not documented as identicalNo direct winnerVersion: 4.0Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Temperature: 1, Top P: 0.95, Output limit: 131,072 tokens	21.212 %	Higher measured value51.515 %	Vals AI Source details Data as of: 10/13/2026
Terminal-Bench ScienceLife Sciences Test details Test setup not documented as identicalNo direct winnerVersion: 0.1Metric: accuracyScoring: Higher is betterMuse Spark 1.3: Reasoning: xhigh, Harness: Mini-SWE-agent, Temperature: 1, Output limit: 131,072 tokens GLM-5.3-Flash: Reasoning: max, Harness: Mini-SWE-agent, Temperature: 1, Top P: 0.95, Output limit: 128,000 tokens	Higher measured value15.789 %	5.263 %	Vals AI Source details Data as of: 10/14/2026
Key to the marks:better valuebetter, but not conclusivetie

Our model tests

Our tests

SVG short test

One SVG prompt and fixed output settings. Captures are shown only with 0 reported reasoning tokens or documented lack of reasoning controls. Identical test conditions do not make the models' internal compute identical.

Shared test round: 4ae5d8eeb9a9

Muse Spark 1.3

Unavailable in this round

GLM-5.3-Flash

Unavailable in this round

Earlier tests

Pelican on a bicycle

A community classic for free-form SVG drawing.

Muse Spark 1.3

Reasoning (requested): high

Current attempt

Run date: Oct 2, 2026
API provider: openrouter
Requested API model: meta/muse-spark-1.3
API model reported in response: meta/muse-spark-1.3
Reasoning setting: high
Pinned endpoint: Meta | meta/muse-spark-1.3-20260902
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 87.9 s

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (2)

Oct 1, 2026Output withheld for safetyReasoning setting: high
Sep 7, 2026API request failedReasoning setting: high

GLM-5.3-Flash

Reasoning (requested): high

Current attempt

Run date: Oct 1, 2026
API provider: openrouter
Requested API model: z-ai/glm-5.3-flash
API model reported in response: z-ai/glm-5.3-flash
Reasoning setting: high
Pinned endpoint: Z.AI | z-ai/glm-5.3-flash-20260826
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 49.7 s

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (1)

Sep 7, 2026API request failedReasoning setting: high

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

How to read this comparison

Winning one benchmark is not an overall verdict

Your workload, budget, and required context length matter most. A coding benchmark says little about visual understanding or agent performance.

The standard price table uses direct developer rates. The provider comparison labels OpenRouter endpoints, regions, and special tariffs separately.

Every result links to its measurement source and retrieval date. Labels distinguish vendor reports, official benchmark leaderboards, and independent evaluations. Disputed or archived results are excluded.

More matchups

Compare Muse Spark 1.3 and GLM-5.3-Flash with other leading models using the same data and benchmark logic.

View the complete LLM comparison

Muse Spark 1.3 vs. GLM-5.3-Flash

Change model selection

General specifications

API pricing compared

Shared benchmarks

Capability profile from matched benchmarks

Our tests

SVG short test

Muse Spark 1.3

GLM-5.3-Flash

Pelican on a bicycle

Winning one benchmark is not an overall verdict

Related LLM comparisons