Model comparison

GPT-5.4 mini vs Qwen3.5-35B-A3B

Data verified July 23, 2026

Head-to-head evidence from 15 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

GPT-5.4 mini

OpenAI

56.77/100

Margin

0.2pts

winning →

Qwen3.5-35B-A3B

Alibaba

56.97/100

1 category wins1 category wins

Public leaderboard positions: GPT-5.4 mini #75 (Estimated); Qwen3.5-35B-A3B #72 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. GPT-5.4 mini and Qwen3.5-35B-A3B share 15 comparable benchmark results. 2 of 8 categories are comparable. 15 results are unique to GPT-5.4 mini; 13 to Qwen3.5-35B-A3B.

Updated July 23, 2026

Shared results: 15
GPT-5.4 mini only: 15
Qwen3.5-35B-A3B only: 13
Comparable categories: 2 / 8

Pick Qwen3.5-35B-A3B if you want the stronger benchmark profile. GPT-5.4 mini only becomes the better choice if agentic is the priority or you need the larger 400K context window.

Confidence note. This is a partial-evidence comparison with 15 shared benchmark results across 6 evidence categories; 2 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Qwen3.5-35B-A3B has the cleaner BenchAlign overall profile here, landing at 56.97 versus 56.77. It is a real lead, but still close enough that category-level strengths matter more than the headline number.

Qwen3.5-35B-A3B's sharpest advantage is in knowledge, where it averages 81.6 against 47.8. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 60% to 40.5%. GPT-5.4 mini does hit back in agentic, so the answer changes if that is the part of the workload you care about most.

GPT-5.4 mini is also the more expensive model on tokens at $0.75 input / $4.50 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Qwen3.5-35B-A3B. That is roughly Infinityx on output cost alone. GPT-5.4 mini gives you the larger context window at 400K, compared with 262K for Qwen3.5-35B-A3B.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for GPT-5.4 mini and Qwen3.5-35B-A3B
Category	GPT-5.4 mini	Δ	Qwen3.5-35B-A3B
Knowledge	GPT-5.4 mini47.8	Margin→ 33.8	Qwen3.5-35B-A3B81.6
Agentic	GPT-5.4 mini65.7	Margin← 14.7	Qwen3.5-35B-A3B51.0
Coding	GPT-5.4 miniNot measured	MarginNo overlap	Qwen3.5-35B-A3B60.6
Reasoning	GPT-5.4 miniNot measured	MarginNo overlap	Qwen3.5-35B-A3B59.0
Math	GPT-5.4 mini21.7	MarginNo overlap	Qwen3.5-35B-A3BNot measured
Multilingual	GPT-5.4 miniNot measured	MarginNo overlap	Qwen3.5-35B-A3B81.0
Multimodal	GPT-5.4 mini76.6	MarginNo overlap	Qwen3.5-35B-A3BNot measured
Inst. Following	GPT-5.4 miniNot measured	MarginNo overlap	Qwen3.5-35B-A3B91.9

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

A · GPT-5.4 miniB · Qwen3.5-35B-A3B

Terminal-Bench 2.0
Agentic
Source ↗
A 60%B 40.5%
Winner: GPT-5.4 miniΔ 19.5
Terminal-Bench 2.0: GPT-5.4 mini scored 60%; Qwen3.5-35B-A3B scored 40.5%. GPT-5.4 mini wins this benchmark.
OSWorld-Verified
Agentic
Source ↗
A 72.1%B 54.5%
Winner: GPT-5.4 miniΔ 17.6
OSWorld-Verified: GPT-5.4 mini scored 72.1%; Qwen3.5-35B-A3B scored 54.5%. GPT-5.4 mini wins this benchmark.
GPQA
Knowledge
Source ↗
A 88%B 84.2%
Winner: GPT-5.4 miniΔ 3.8
GPQA: GPT-5.4 mini scored 88%; Qwen3.5-35B-A3B scored 84.2%. GPT-5.4 mini wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

Metric	GPT-5.4 mini	Qwen3.5-35B-A3B	Comparison
Input / output priceUSD per 1M tokens	GPT-5.4 mini$0.75 input / $4.5 output	Qwen3.5-35B-A3B$0 input / $0 output	Qwen3.5-35B-A3B has the lower combined listed price.
Generation speedtokens per second	GPT-5.4 mini201 tok/s	Qwen3.5-35B-A3BNot available	A complete speed comparison is not available.
First-answer latencyseconds to first token	GPT-5.4 mini3.85 s	Qwen3.5-35B-A3BNot available	A complete latency comparison is not available.
Context windowmaximum listed tokens	GPT-5.4 mini400K	Qwen3.5-35B-A3B262K	GPT-5.4 mini lists the larger context window.

Benchmark Deep Dive

AgenticGPT-5.4 mini wins

11 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
Terminal-Bench 2.0Source	60%	40.5%	GPT-5.4 mini leads
OSWorld-VerifiedSource	72.1%	54.5%	GPT-5.4 mini leads
MCP AtlasSource	57.7%	—	Not comparable
ToolathlonSource	42.9%	—	Not comparable
τ²-bench resultsSource	83.3%	89.2%	Qwen3.5-35B-A3B leads
AA Agentic IndexSource	30.2%	—	Not comparable
APEX-Agents-AASource	28.2%	—	Not comparable
GDPval-AASource	33.6%	—	Not comparable
GDPval-AASource	1171	—	Not comparable
BrowseCompSource	—	61%	Not comparable
Gert LabsSource	—	28.96%	Not comparable

Coding

6 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
Vibe Code BenchSource	47.97%	—	Not comparable
AA Coding IndexSource	56.1%	—	Not comparable
AA-SciCodeSource	49.9%	37.7%	GPT-5.4 mini leads
FrontierCode 1.1 MainSource	27.0%	—	Not comparable
SWE-bench VerifiedSource	—	69.2%	Not comparable
SWE-RebenchSource	—	53.7%	Not comparable

Reasoning

3 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
AA-LCRSource	69.3%	62.7%	GPT-5.4 mini leads
CritPtSource	10.0%	0.9%	GPT-5.4 mini leads
LongBench v2Source	—	59%	Not comparable

KnowledgeQwen3.5-35B-A3B wins

11 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
GPQASource	88%	84.2%	GPT-5.4 mini leads
HLESource	41.5%	—	Not comparable
HLE w/o toolsSource	28.2%	—	Not comparable
Artificial Analysis Intelligence IndexSource	40.0%	29.3%	GPT-5.4 mini leads
AA-GPQA DiamondSource	87.5%	84.5%	GPT-5.4 mini leads
AA-HLESource	26.6%	19.7%	GPT-5.4 mini leads
AA-Omniscience IndexSource	-18.7%	-46.4%	GPT-5.4 mini leads
AA-Omniscience AccuracySource	37.5%	20.5%	GPT-5.4 mini leads
AA-Omniscience Hallucination RateSource	89.8%	84.0%	Qwen3.5-35B-A3B leads
MMLU-ProSource	—	85.3%	Not comparable
SuperGPQASource	—	63.4%	Not comparable

Math

2 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
FrontierMath v2 (Tiers 1-3)Source	28.280%	—	Not comparable
FrontierMath v2 (Tier 4)Source	2.080%	—	Not comparable

Multilingual

1 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
MMLU-ProXSource	—	81%	Not comparable

Multimodal

7 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
MMMU-ProSource	76.6%	—	Not comparable
MMMU-Pro w/ PythonSource	78%	—	Not comparable
AA-MMMU-ProSource	73.3%	72.7%	GPT-5.4 mini leads
MMMUSource	—	81.4%	Not comparable
MMVUSource	—	72.3%	Not comparable
MathVisionSource	—	83.9%	Not comparable
V*Source	—	92.7%	Not comparable

Inst. Following

2 benchmarks

Benchmark	GPT-5.4 mini	Qwen3.5-35B-A3B	Result
AA-IFBenchSource	73.3%	72.5%	GPT-5.4 mini leads
IFEvalSource	—	91.9%	Not comparable

Frequently Asked Questions (3)

Which is better, GPT-5.4 mini or Qwen3.5-35B-A3B?

Qwen3.5-35B-A3B is ahead on BenchLM's BenchAlign leaderboard, 56.97 to 56.77. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 60% and 40.5%.

Which is better for knowledge tasks, GPT-5.4 mini or Qwen3.5-35B-A3B?

Qwen3.5-35B-A3B has the edge for knowledge tasks in this comparison, averaging 81.6 versus 47.8. Inside this category, AA-Omniscience Index is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, GPT-5.4 mini or Qwen3.5-35B-A3B?

GPT-5.4 mini has the edge for agentic tasks in this comparison, averaging 65.7 versus 51. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.

Related Comparisons

Explore More

OpenAI Compare Pricing Methodology Find Your Best LLM Overall Rankings

Last updated: July 23, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.