Model comparison

Qwen3.8 Max vs Sonar

Qwen3.8 Max is the stronger model overall, scoring 56.8 to 38.5 on the Noometry Index. Sonar costs 3.0× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.

Last verified October 10, 2026. 0 shared benchmarks.

Qwen3.8 Max

Alibaba (Qwen)

56.8

Rank #22 Confirmed

Sonar

Perplexity

38.5

Rank #187 Confirmed

Summary

The widest gap is in math, where Qwen3.8 Max leads 73.2 to 33.7.
Sonar is cheaper at $1 / $1 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
Qwen3.8 Max accepts more context: 1M tokens versus 128K.

Side by side

Qwen3.8 Max and Sonar specifications
	Qwen3.8 Max	Sonar
Provider	Alibaba (Qwen)	Perplexity
Noometry Index	56.8	38.5
Released	2026-08-02	2024-01-01
Weights	Proprietary	Proprietary
Context window	1M	128K
Max output	131K	4K
Input $ / M tokens	$2	$1
Output $ / M tokens	$6	$1
Results tracked	39	7

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Qwen3.8 Max leads

Qwen3.8 Max: 53.5 (#29), Sonar: 35.7 (#221)

Coding benchmarks
Benchmark	Qwen3.8 Max	Sonar
DeepSWE	57.5%	—
LMArena WebDev	1674	—
FrontierSWE	17.8%	—
SciCode	53.2%	—
LiveBench Coding	—	35.1%
LMArena Coding	1502	—

Agentic & Tool Use Not comparable

Qwen3.8 Max: 45.4 (#14), Sonar: —

Agentic & Tool Use benchmarks
Benchmark	Qwen3.8 Max	Sonar
APEX-Agents	63.3%	—
τ²-bench Banking	55.1%	—
GDP.pdf	23.2%	—

Reasoning Qwen3.8 Max leads

Qwen3.8 Max: 54.4 (#26), Sonar: 21.1 (#227)

Reasoning benchmarks
Benchmark	Qwen3.8 Max	Sonar
NYT Connections (extended)	88.3%	—
CritPt	20%	—
Chess Puzzles	40%	—
LiveBench Reasoning	—	46.3%
LMArena Hard Prompts	1496	—
Mystery Game Puzzles	38%	—
DTBench	92%	—
LiveBench Data Analysis	—	37.9%
LMCA	46.2%	—
Epoch Capabilities Index	156.41	—
LiveBench	—	46.9%

Math Qwen3.8 Max leads

Qwen3.8 Max: 73.2 (#20), Sonar: 33.7 (#200)

Math benchmarks
Benchmark	Qwen3.8 Max	Sonar
FrontierMath (Tiers 1-3)	74.7%	—
FrontierMath Tier 4	46.3%	—
OTIS Mock AIME 2024-2025	100%	—
ProofBench	58%	—
LiveBench Math	—	41.6%
LMArena Math	1499	—

Knowledge Not comparable

Qwen3.8 Max: 61.7 (#27), Sonar: —

Knowledge benchmarks
Benchmark	Qwen3.8 Max	Sonar
GPQA Diamond	92.7%	—
SimpleQA Verified	47.3%	—
LMArena Expert	1507	—

Multimodal Not comparable

Qwen3.8 Max: 37.2 (#75), Sonar: —

Multimodal benchmarks
Benchmark	Qwen3.8 Max	Sonar
LMArena Vision	1314	—
Furniture Assembly	20%	—

Multilingual Not comparable

Qwen3.8 Max: 56.7 (#18), Sonar: —

Multilingual benchmarks
Benchmark	Qwen3.8 Max	Sonar
LMArena Non-English	1472	—
LMArena Chinese	1538	—
LMArena French	1503	—
LMArena German	1483	—
LMArena Japanese	1467	—
LMArena Korean	1461	—
LMArena Russian	1481	—
LMArena Spanish	1492	—

Instruction Following Qwen3.8 Max leads

Qwen3.8 Max: 77.6 (#17), Sonar: 71.4 (#150)

Instruction Following benchmarks
Benchmark	Qwen3.8 Max	Sonar
LiveBench Instruction Following	—	76.2%
LMArena Instruction Following	1479	—

Long Context Not comparable

Qwen3.8 Max: 45.6 (#31), Sonar: —

Long Context benchmarks
Benchmark	Qwen3.8 Max	Sonar
LMArena Longer Query	1489	—

Writing & Preference Qwen3.8 Max leads

Qwen3.8 Max: 67.1 (#30), Sonar: 52.6 (#167)

Writing & Preference benchmarks
Benchmark	Qwen3.8 Max	Sonar
LMArena Text	1483	—
LMArena Creative Writing	1479	—
LMArena Multi-Turn	1489	—
LiveBench Language	—	44.1%

Frequently asked questions

Is Qwen3.8 Max better than Sonar?

Which is cheaper, Qwen3.8 Max or Sonar?

Sonar is cheaper. It lists at $1 per million input tokens and $1 per million output tokens; Qwen3.8 Max lists at $2 and $6.

Is Qwen3.8 Max or Sonar better for coding?

Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 35.7 in the Noometry coding category.

Which has the bigger context window?

Qwen3.8 Max does, with 1M tokens against 128K.

How many benchmarks do Qwen3.8 Max and Sonar share?

0 benchmarks have published results for both models. Qwen3.8 Max has 39 scored results on Noometry and Sonar has 7.