Model comparison

Phi-4 vs Qwen3.5 122B-A10B

Qwen3.5 122B-A10B is the stronger model overall, scoring 42.1 to 31.2 on the Noometry Index. Phi-4 costs 13× less per token, which makes it the better buy when Qwen3.5 122B-A10B's lead doesn't matter for your workload.

Last verified October 12, 2026. 18 shared benchmarks.

Phi-4

Microsoft

31.2

Rank #279 Confirmed

Qwen3.5 122B-A10B

Alibaba (Qwen)

42.1

Rank #119 Confirmed

Summary

They share 18 benchmarks with published results for both. Phi-4 scores higher in 0 categories and Qwen3.5 122B-A10B in 8 categories; 8 gaps are clear of the uncertainty.
The widest gap is in writing & preference, where Qwen3.5 122B-A10B leads 60.0 to 40.5.
The biggest single-benchmark swing is Vectara Hallucination Rate: 3.7% for Phi-4 and 11.2% for Qwen3.5 122B-A10B.
Phi-4 is cheaper at $0.07 / $0.14 per million input/output tokens, against $0.40 / $3.20 for Qwen3.5 122B-A10B.
Qwen3.5 122B-A10B accepts more context: 262K tokens versus 128K.

Side by side

Phi-4 and Qwen3.5 122B-A10B specifications
	Phi-4	Qwen3.5 122B-A10B
Provider	Microsoft	Alibaba (Qwen)
Noometry Index	31.2	42.1
Released	2024-12-11	2026-02-23
Weights	Open	Open
Context window	128K	262K
Max output	4K	66K
Input $ / M tokens	$0.07	$0.40
Output $ / M tokens	$0.14	$3.20
Results tracked	37	27

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

Category by category

Coding Qwen3.5 122B-A10B leads

Phi-4: 34.4 (#239), Qwen3.5 122B-A10B: 39.1 (#162)

Coding benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Coding	1231	1436
LMArena WebDev	—	1360
SciCode	—	35.6%
BigCodeBench Instruct	45.5%	—
LiveBench Coding	30.7%	—
BigCodeBench Complete	55.4%	—

Agentic & Tool Use Not comparable

Phi-4: 22.8 (#128), Qwen3.5 122B-A10B: —

Agentic & Tool Use benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
Berkeley Function Calling Leaderboard	28.8%	—
BALROG	11.6%	—

Reasoning Qwen3.5 122B-A10B leads

Phi-4: 17.7 (#291), Qwen3.5 122B-A10B: 27.2 (#123)

Reasoning benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Hard Prompts	1220	1421
NYT Connections (extended)	—	51.7%
CritPt	—	0.9%
Chess Puzzles	1%	—
Thematic Generalization	—	51.2%
LiveBench Reasoning	47.8%	—
Mystery Game Puzzles	—	17%
DTBench	—	84.3%
LiveBench Data Analysis	45.2%	—
LMCA	—	32.2%
Epoch Capabilities Index	130.42	—
LiveBench	41.6%	—

Math Qwen3.5 122B-A10B leads

Phi-4: 20.8 (#285), Qwen3.5 122B-A10B: 39.1 (#112)

Math benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Math	1246	1432
OTIS Mock AIME 2024-2025	13.8%	—
LiveBench Math	42%	—
MATH Level 5	64.9%	—

Knowledge Qwen3.5 122B-A10B leads

Phi-4: 32.6 (#209), Qwen3.5 122B-A10B: 38.8 (#142)

Knowledge benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
Vectara Hallucination Rate	3.7%	11.2%
LMArena Expert	1203	1432
GPQA Diamond	56.1%	—
Confabulations	29.4%	—
MMLU	84.8%	—

Multimodal Not comparable

Phi-4: —, Qwen3.5 122B-A10B: 39.6 (#57)

Multimodal benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Vision	—	1245

Multilingual Qwen3.5 122B-A10B leads

Phi-4: 37.2 (#237), Qwen3.5 122B-A10B: 51.6 (#107)

Multilingual benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Non-English	1197	1400
LMArena Chinese	1212	1462
LMArena French	1224	1442
LMArena German	1222	1426
LMArena Japanese	1158	1367
LMArena Korean	1151	1352
LMArena Russian	1209	1400
LMArena Spanish	1234	1424

Instruction Following Qwen3.5 122B-A10B leads

Phi-4: 60.4 (#251), Qwen3.5 122B-A10B: 73.8 (#115)

Instruction Following benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Instruction Following	1201	1399
LiveBench Instruction Following	58.4%	—

Long Context Qwen3.5 122B-A10B leads

Phi-4: 36.9 (#226), Qwen3.5 122B-A10B: 43.0 (#109)

Long Context benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Longer Query	1217	1410

Writing & Preference Qwen3.5 122B-A10B leads

Phi-4: 40.5 (#244), Qwen3.5 122B-A10B: 60.0 (#105)

Writing & Preference benchmarks
Benchmark	Phi-4	Qwen3.5 122B-A10B
LMArena Text	1217	1417
LMArena Creative Writing	1182	1368
LMArena Multi-Turn	1206	1416
Short-Story Creative Writing	62.6%	—
LiveBench Language	25.6%	—

Frequently asked questions

Is Phi-4 better than Qwen3.5 122B-A10B?

Which is cheaper, Phi-4 or Qwen3.5 122B-A10B?

Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Qwen3.5 122B-A10B lists at $0.40 and $3.20.

Is Phi-4 or Qwen3.5 122B-A10B better for coding?

Qwen3.5 122B-A10B scores higher on coding benchmarks: 39.1 versus 34.4 in the Noometry coding category.

Which has the bigger context window?

Qwen3.5 122B-A10B does, with 262K tokens against 128K.

How many benchmarks do Phi-4 and Qwen3.5 122B-A10B share?

18 benchmarks have published results for both models. Phi-4 has 37 scored results on Noometry and Qwen3.5 122B-A10B has 27.