Muse Spark 1.3 vs Qwen3.8 Max

At a glance

• Capability index

  • Muse Spark 1.3156.8
  • Qwen3.8 Max156.4

  Muse Spark 1.3 is 0.3 points higher.

• Price per 1M tokens

  • Muse Spark 1.3$2.00
  • Qwen3.8 Max$3.00

  Muse Spark 1.3 is 33% cheaper.

• Context window

  • Muse Spark 1.31M tokens
  • Qwen3.8 Max1M tokens

  They take in about the same.

• Released

  • Muse Spark 1.3Sep 2, 2026
  • Qwen3.8 MaxAug 3, 2026

  Muse Spark 1.3 is 30 days newer.

Where each scores higher

On the 19 tests both have taken, each on its own scale. A gap of a point or two can sit within a test’s margin of error.

Level on FrontierMath Tier 4.

4 more tests have results for only one of them; the table below lists every result.

Everything side by side

Comparison of Muse Spark 1.3 vs Qwen3.8 Max

Measure	Muse Spark 1.3Meta AI	Qwen3.8 MaxAlibaba
Model
Maker	Meta AI	Alibaba
Sold here by	DevPass (LLM Gateway)	Alibaba
API model ID	muse-spark-1.3	qwen3.8-max
Released	Sep 2, 2026	Aug 3, 2026
Knowledge cutoff	Not reported	Not reported
Weights	Closed: hosted access only	Closed: hosted access only
Developer’s country	United States	China
Hosts selling it	11	36including Alibaba directly
Test results
Capability index	156.8 (best of these models)#19 of 274	156.4#22 of 274
LiveBench overall	81.6% (best of these models)#7 of 66	78.5%#17 of 66
Arena text	1494 (best of these models)#9 of 413	1482#23 of 413
LiveBench coding	81.1% (best of these models)#17 of 66	72.9%#52 of 66
Arena WebDev	1657#14 of 138	1671 (best of these models)#10 of 138
Arena coding	1539 (best of these models)#11 of 408	1524#28 of 408
LiveBench mathematics	96.0% (best of these models)#12 of 66	91.3%#27 of 66
Mock AIME 2024–2025	99.2%#15 of 297	99.4% (best of these models)#14 of 297
FrontierMath Tiers 1–3	74.4%#19 of 114	74.7% (best of these models)#18 of 114
FrontierMath Tier 4	46.3%#25 of 70	46.3%#25 of 70
Arena math	1509 (best of these models)#9 of 396	1498#18 of 396
SimpleQA Verified	Not tested	45.8%#37 of 86
LiveBench agentic coding	64.1%#9 of 66	64.7% (best of these models)#7 of 66
CursorBench	41.6%#22 of 62	Not tested
DeepSWE	Not tested	57.5%#36 of 69
LiveBench language	82.8% (best of these models)#24 of 66	79.7%#34 of 66
LiveBench instructions	78.0% (best of these models)#4 of 66	74.1%#13 of 66
Arena creative writing	1459#28 of 411	1468 (best of these models)#20 of 411
Arena instruction following	1486 (best of these models)#16 of 413	1474#31 of 413
LiveBench data analysis	79.6% (best of these models)#12 of 66	78.4%#24 of 66
LiveBench reasoning	89.7% (best of these models)#15 of 66	88.2%#23 of 66
Chess puzzles	38.0% (best of these models)#32 of 227	29.0%#54 of 227
Arena hard prompts	1517 (best of these models)#10 of 413	1503#27 of 413
GPQA Diamond	Not tested	92.7%#24 of 319
Price per million tokens
Input	$1.25 (best of these models)	$2.00
Output	$4.25 (best of these models)	$6.00
Blended, 3 input to 1 output	$2.00 (best of these models)	$3.00
Cached input	$0.15 (best of these models)	$0.25
Writing to the cache	Not reported	$2.50
The maker’s own price	Not sold directly	This listing
Limits
Context window	1M tokens	1M tokens
Max input	1M tokens (best of these models)	991K tokens
Max output	131.1K tokens	131.1K tokens
Features
Reads	audio, images, PDFs, text and video	images, PDFs, text and video
Produces	text	text
Reasoning	Yeseffort minimal, low, medium, high, xhigh	Yeseffort low, medium, xhigh; can be switched off; thinking budget can be set
Tool calling	Yes	Yes
Structured output	Yes, such as JSON	Yes, such as JSON
File attachments	Yes	Yes
Temperature setting	Supported	Supported

Bold marks the better value in each row. Prices are each model’s own list price, or the middle price across hosts when the maker does not sell it, blended as three input tokens for every output token. Test results as published by LiveBench (Apache-2.0), Epoch AI (CC BY 4.0), Datacurve DeepSWE, compiled by Epoch AI (CC BY 4.0), Cursor CursorBench, compiled by Epoch AI (CC BY 4.0), Arena (CC BY 4.0); model details and prices from models.dev (MIT). Confirm pricing with the provider before use.