Model A

Muse Spark 1.1 vs Qwen3.5 397B

Updated August 22, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Model B

Qwen3.5 397B

Alibaba

57.5/100

Estimated · Public rank #84

90% interval 46.0–69.0

Decision reading

Muse Spark 1.1 has the higher public score, 76.74 versus 57.49, and the 90% score intervals do not overlap.

6 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

Long documents
Prompts that approach the documented context limit
Muse Spark 1.1
Muse Spark 1.1 has the larger documented context window.
Confidence: documented

Show secondary and unsupported calls

Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Qwen3.5 397B does not fit this workload in one request. Qwen3.5 397B has no published cached-input rate, so cached tokens use its listed input rate. Muse Spark 1.1 has no comparable published API token rate.
Confidence: rate-fallback
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

15 Muse Spark 1.1 only

6 Shared

32 Qwen3.5 397B only

Shared results: 6
Muse Spark 1.1 only: 15
Qwen3.5 397B only: 32
Like-for-like categories: 0 / 8

4 categories use different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Category results, on a stated basis

Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.

Agentic

Directional only

Muse Spark 1.1: 80.4
Qwen3.5 397B: 56.5
Weighted basis: 2 vs 2 rows
Reading: Directional only

Coding

Directional only

Muse Spark 1.1: 61.5
Qwen3.5 397B: 66.5
Weighted basis: 1 vs 2 rows
Reading: Directional only

Knowledge

Directional only

Muse Spark 1.1: 62.1
Qwen3.5 397B: 56.6
Weighted basis: 1 vs 4 rows
Reading: Directional only

Multimodal

Directional only

Muse Spark 1.1: 88.4
Qwen3.5 397B: 79.6
Weighted basis: 1 vs 2 rows
Reading: Directional only

Reasoning

Not comparable

Muse Spark 1.1: Not measured
Qwen3.5 397B: 63.2
Weighted basis: 0 vs 1 rows
Reading: Not comparable

Math

Not comparable

Muse Spark 1.1: Not measured
Qwen3.5 397B: 90.6
Weighted basis: 0 vs 2 rows
Reading: Not comparable

Multilingual

Not comparable

Muse Spark 1.1: Not measured
Qwen3.5 397B: 84.7
Weighted basis: 0 vs 1 rows
Reading: Not comparable

Instruction following

Not comparable

Muse Spark 1.1: Not measured
Qwen3.5 397B: 92.6
Weighted basis: 0 vs 1 rows
Reading: Not comparable

Category averages with the server-provided evidence basis for Muse Spark 1.1 and Qwen3.5 397B
Category	Muse Spark 1.1	Qwen3.5 397B	Weighted basis	Reading
Agentic	80.4	56.5	Directional only2 vs 2 rows	Directional only
Coding	61.5	66.5	Directional only1 vs 2 rows	Directional only
Knowledge	62.1	56.6	Directional only1 vs 4 rows	Directional only
Multimodal	88.4	79.6	Directional only1 vs 2 rows	Directional only
Reasoning	Not measured	63.2	Not comparable0 vs 1 rows	Not comparable
Math	Not measured	90.6	Not comparable0 vs 2 rows	Not comparable
Multilingual	Not measured	84.7	Not comparable0 vs 1 rows	Not comparable
Instruction following	Not measured	92.6	Not comparable0 vs 1 rows	Not comparable

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

HLE
Knowledge
Muse Spark 1.1: 62.1%Qwen3.5 397B: 28.7%Normalized gap 33.4Muse Spark 1.1 source Qwen3.5 397B source
Terminal-Bench 2.0
Agentic
Muse Spark 1.1: 80%Qwen3.5 397B: 52.5%Normalized gap 27.5Muse Spark 1.1 source Qwen3.5 397B source
SWE-bench Pro
Coding
Muse Spark 1.1: 61.5%Qwen3.5 397B: 50.9%Normalized gap 10.6Muse Spark 1.1 source Qwen3.5 397B source
CharXiv
Multimodal
Muse Spark 1.1: 88.4%Qwen3.5 397B: 80.8%Normalized gap 7.6Muse Spark 1.1 source Qwen3.5 397B source

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

Muse Spark 1.1: API rate not published; Fits in one request
Qwen3.5 397B: $0.0024; Fits in one request

Muse Spark 1.1 has no comparable published API token rate.

Repository review

50K fresh input + 3K output tokens

Muse Spark 1.1: API rate not published; Fits in one request
Qwen3.5 397B: $0.0408; Fits in one request

Muse Spark 1.1 has no comparable published API token rate.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Muse Spark 1.1: API rate not published; Fits in one request; Cached-input rate unavailable
Qwen3.5 397B: $0.168; Does not fit in one request; Cached input priced at the published list-input rate

Qwen3.5 397B does not fit this workload in one request. Qwen3.5 397B has no published cached-input rate, so cached tokens use its listed input rate. Muse Spark 1.1 has no comparable published API token rate.

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

SpecificationMuse Spark 1.1Qwen3.5 397B

Context window

Maximum documented context; output-token limits may be lower.

Muse Spark 1.1

Qwen3.5 397B

128K

API model ID

Muse Spark 1.1

Not sourced

Qwen3.5 397B

Not sourced

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

Muse Spark 1.1

No comparable hosted API rate

Qwen3.5 397B

Not published

Documented inputs

Muse Spark 1.1

Not sourced

Qwen3.5 397B

Not sourced

Documented outputs

Muse Spark 1.1

Not sourced

Qwen3.5 397B

Not sourced

Provider availability

Muse Spark 1.1

Not sourced

Qwen3.5 397B

Not sourced

Reasoning profile

Muse Spark 1.1

Reasoning

Qwen3.5 397B

Non-Reasoning

Weight access

Muse Spark 1.1

Proprietary

Qwen3.5 397B

Open Weight

License

Muse Spark 1.1

Proprietary

Qwen3.5 397B

Open Weight

Release date

Muse Spark 1.1

2026-07-09

Qwen3.5 397B

2026-02-16

If you already use one of these models

Deployment change: The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal: Muse Spark 1.1 has the higher public score, 76.74 versus 57.49, and the 90% score intervals do not overlap.
Workload cost: A complete comparable API-rate estimate is not available for both models.
Context tradeoff: Muse Spark 1.1 has the larger documented window (1M).

Run the same representative tasks against both endpoints before changing production traffic.

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence53 rows

Agentic

Terminal-Bench 2.0
Muse Spark 1.180%
Source
Qwen3.5 397B52.5%
Source
Muse Spark 1.1 leads this result
MCP Atlas
Muse Spark 1.188.1%
Source
Qwen3.5 397B46.1%
Source
Muse Spark 1.1 leads this result
Toolathlon
Muse Spark 1.175.6%
Source
Qwen3.5 397B36.3%
Source
Muse Spark 1.1 leads this result
OSWorld-Verified
Muse Spark 1.180.8%
Source
Qwen3.5 397B—
Not directly comparable
WebArena-Verified
Muse Spark 1.169%
Source
Qwen3.5 397B—
Not directly comparable
DeepSearchQA
Muse Spark 1.184.9%
Source
Qwen3.5 397B—
Not directly comparable
CyberGym
Muse Spark 1.159.0%
Source
Qwen3.5 397B—
Not directly comparable
Finance Agent v2
Muse Spark 1.157.2%
Source
Qwen3.5 397B—
Not directly comparable
deepSwe
Muse Spark 1.153.3%
Source
Qwen3.5 397B—
Not directly comparable
OSWorld 2.0
Muse Spark 1.114.2%
Source
Qwen3.5 397B—
Not directly comparable
JobBench
Muse Spark 1.154.7%
Source
Qwen3.5 397B—
Not directly comparable
Cybench
Muse Spark 1.192.9%
Source
Qwen3.5 397B—
Not directly comparable
ExploitGym
Muse Spark 1.10.8%
Source
Qwen3.5 397B—
Not directly comparable
BrowseComp
Muse Spark 1.1—
Qwen3.5 397B62%
Source
Not directly comparable
Claw-Eval
Muse Spark 1.1—
Qwen3.5 397B56.8%
Source
Not directly comparable
QwenClawBench
Muse Spark 1.1—
Qwen3.5 397B51.8%
Source
Not directly comparable
τ³-bench results
Muse Spark 1.1—
Qwen3.5 397B68.4%
Source
Not directly comparable
VITA-Bench
Muse Spark 1.1—
Qwen3.5 397B43.7%
Source
Not directly comparable
DeepPlanning
Muse Spark 1.1—
Qwen3.5 397B37.6%
Source
Not directly comparable
MCP-Tasks
Muse Spark 1.1—
Qwen3.5 397B74.2%
Source
Not directly comparable
WideResearch
Muse Spark 1.1—
Qwen3.5 397B74.0%
Source
Not directly comparable
Gert Labs
Muse Spark 1.1—
Qwen3.5 397B46.76%
Source
Not directly comparable
ResearchClawBench
Muse Spark 1.1—
Qwen3.5 397B14.2%
Source
Not directly comparable

Coding

Terminal-Bench 2.0
Muse Spark 1.180.0%
Source
Qwen3.5 397B—
Not directly comparable
SWE-bench Pro
Muse Spark 1.161.5%
Source
Qwen3.5 397B50.9%
Source
Muse Spark 1.1 leads this result
SWE-bench Verified
Muse Spark 1.1—
Qwen3.5 397B76.2%
Source
Not directly comparable
LiveCodeBench v6
Muse Spark 1.1—
Qwen3.5 397B83.6%
Source
Not directly comparable

Reasoning

MRCR 1M
Muse Spark 1.154.1%
Source
Qwen3.5 397B—
Not directly comparable
LongBench v2
Muse Spark 1.1—
Qwen3.5 397B63.2%
Source
Not directly comparable
AI-Needle
Muse Spark 1.1—
Qwen3.5 397B68.7%
Source
Not directly comparable

Knowledge

HLE
Muse Spark 1.162.1%
Source
Qwen3.5 397B28.7%
Source
Muse Spark 1.1 leads this result
HLE w/o tools
Muse Spark 1.152.2%
Source
Qwen3.5 397B—
Not directly comparable
HealthBench Professional
Muse Spark 1.159.3%
Source
Qwen3.5 397B—
Not directly comparable
GPQA
Muse Spark 1.1—
Qwen3.5 397B88.4%
Source
Not directly comparable
SuperGPQA
Muse Spark 1.1—
Qwen3.5 397B70.4%
Source
Not directly comparable
MMLU-Pro
Muse Spark 1.1—
Qwen3.5 397B87.8%
Source
Not directly comparable
MMLU-Redux
Muse Spark 1.1—
Qwen3.5 397B94.9%
Source
Not directly comparable
C-Eval
Muse Spark 1.1—
Qwen3.5 397B93%
Source
Not directly comparable

Math

AIME26
Muse Spark 1.1—
Qwen3.5 397B93.3%
Source
Not directly comparable
HMMT Feb 2025
Muse Spark 1.1—
Qwen3.5 397B94.8%
Source
Not directly comparable
HMMT Nov 2025
Muse Spark 1.1—
Qwen3.5 397B92.7%
Source
Not directly comparable
HMMT Feb 2026
Muse Spark 1.1—
Qwen3.5 397B87.9%
Source
Not directly comparable
MMAnswerBench
Muse Spark 1.1—
Qwen3.5 397B80.9%
Source
Not directly comparable

Multilingual

MMLU-ProX
Muse Spark 1.1—
Qwen3.5 397B84.7%
Source
Not directly comparable
NOVA-63
Muse Spark 1.1—
Qwen3.5 397B59.1%
Source
Not directly comparable

Multimodal

CharXiv
Muse Spark 1.188.4%
Source
Qwen3.5 397B80.8%
Source
Muse Spark 1.1 leads this result
BabyVision
Muse Spark 1.176.3%
Source
Qwen3.5 397B—
Not directly comparable
MMMU-Pro
Muse Spark 1.1—
Qwen3.5 397B79%
Source
Not directly comparable
MathVision
Muse Spark 1.1—
Qwen3.5 397B88.6%
Source
Not directly comparable
VideoMMMU
Muse Spark 1.1—
Qwen3.5 397B84.7%
Source
Not directly comparable
ScreenSpot Pro
Muse Spark 1.1—
Qwen3.5 397B65.6%
Source
Not directly comparable
V*
Muse Spark 1.1—
Qwen3.5 397B95.8%
Source
Not directly comparable

Instruction following

IFEval
Muse Spark 1.1—
Qwen3.5 397B92.6%
Source
Not directly comparable

Frequently asked questions

Which is better, Muse Spark 1.1 or Qwen3.5 397B?

Muse Spark 1.1 has the higher public score, 76.74 versus 57.49, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.

Which is better for coding, Muse Spark 1.1 or Qwen3.5 397B?

The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, Muse Spark 1.1 or Qwen3.5 397B?

The current agentic tasks averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.

Which costs less, Muse Spark 1.1 or Qwen3.5 397B?

Both models do not have comparable published API token rates, so this page does not name a universal price winner.

Which has the larger context window, Muse Spark 1.1 or Qwen3.5 397B?

Muse Spark 1.1 has the larger documented context window: 1M, compared with 128K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated August 22, 2026

Watch Muse Spark 1.1 vs Qwen3.5 397B

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.

Muse Spark 1.1 vs Qwen3.5 397B

Which one for your work

Long documents

Coding work

Agentic work

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

What is actually comparable

Category results, on a stated basis

Agentic

Coding

Knowledge

Multimodal

Reasoning

Math

Multilingual

Instruction following

Shape of the matched evidence

What each workload costs

Chat turn

Repository review

Cache-heavy agent loop

Specification differences

Context window

API model ID

Cached-input rate

Documented inputs

Documented outputs

Provider availability

Reasoning profile

Weight access

License

Release date

Benchmark evidence

Agentic

Coding

Reasoning

Knowledge

Math

Multilingual

Multimodal

Instruction following

Frequently asked questions

Which is better, Muse Spark 1.1 or Qwen3.5 397B?

Which is better for coding, Muse Spark 1.1 or Qwen3.5 397B?

Which is better for agentic tasks, Muse Spark 1.1 or Qwen3.5 397B?

Which costs less, Muse Spark 1.1 or Qwen3.5 397B?

Which has the larger context window, Muse Spark 1.1 or Qwen3.5 397B?

Related comparisons

Watch Muse Spark 1.1 vs Qwen3.5 397B