Model A

OpenAI

73.2/100

Supported · Public rank #12

90% interval 70.1–76.2

GPT-5.4 vs Qwen3.5-122B-A10B

Updated August 13, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Model B

Qwen3.5-122B-A10B

Alibaba

59.5/100

Supported · Public rank #61

90% interval 48.3–70.7

Decision reading

GPT-5.4 has the higher public score estimate, 73.16 versus 59.47, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.

5 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.

Agentic work
Tool use, computer use, and multi-step task completion
GPT-5.4
GPT-5.4 leads on the same 3 weighted benchmark rows.
Confidence: stronger
Long documents
Prompts that approach the documented context limit
GPT-5.4
GPT-5.4 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
No shared weighted benchmark basis supports a winner.
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
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
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.

33 GPT-5.4 only

5 Shared

10 Qwen3.5-122B-A10B only

Shared results: 5
GPT-5.4 only: 33
Qwen3.5-122B-A10B only: 10
Like-for-like categories: 1 / 8

2 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

Like-for-like

GPT-5.4: 77.2
Qwen3.5-122B-A10B: 56.4
Weighted basis: 3 vs 3 rows
Reading: GPT-5.4 leads

Knowledge

Directional only

GPT-5.4: 57.6
Qwen3.5-122B-A10B: 83.6
Weighted basis: 2 vs 3 rows
Reading: Directional only

Multimodal

Directional only

GPT-5.4: 73.2
Qwen3.5-122B-A10B: 77.2
Weighted basis: 3 vs 1 rows
Reading: Directional only

Coding

Not comparable

GPT-5.4: 57.7
Qwen3.5-122B-A10B: 72.0
Weighted basis: 1 vs 1 rows
Reading: Not comparable

Reasoning

Not comparable

GPT-5.4: 74.0
Qwen3.5-122B-A10B: 60.2
Weighted basis: 1 vs 1 rows
Reading: Not comparable

Math

Not comparable

GPT-5.4: 42.5
Qwen3.5-122B-A10B: Not measured
Weighted basis: 2 vs 0 rows
Reading: Not comparable

Multilingual

Not comparable

GPT-5.4: Not measured
Qwen3.5-122B-A10B: 82.2
Weighted basis: 0 vs 1 rows
Reading: Not comparable

Instruction following

Not comparable

GPT-5.4: Not measured
Qwen3.5-122B-A10B: 93.4
Weighted basis: 0 vs 1 rows
Reading: Not comparable

Category averages with the server-provided evidence basis for GPT-5.4 and Qwen3.5-122B-A10B
Category	GPT-5.4	Qwen3.5-122B-A10B	Weighted basis	Reading
Agentic	77.2	56.4	Like-for-like3 vs 3 rows	GPT-5.4 leads
Knowledge	57.6	83.6	Directional only2 vs 3 rows	Directional only
Multimodal	73.2	77.2	Directional only3 vs 1 rows	Directional only
Coding	57.7	72.0	Not comparable1 vs 1 rows	Not comparable
Reasoning	74.0	60.2	Not comparable1 vs 1 rows	Not comparable
Math	42.5	Not measured	Not comparable2 vs 0 rows	Not comparable
Multilingual	Not measured	82.2	Not comparable0 vs 1 rows	Not comparable
Instruction following	Not measured	93.4	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.

Terminal-Bench 2.0
Agentic
GPT-5.4: 75.1%Qwen3.5-122B-A10B: 49.4%Normalized gap 25.7GPT-5.4 source Qwen3.5-122B-A10B source
BrowseComp
Agentic
GPT-5.4: 82.7%Qwen3.5-122B-A10B: 63.8%Normalized gap 18.9GPT-5.4 source Qwen3.5-122B-A10B source
OSWorld-Verified
Agentic
GPT-5.4: 75%Qwen3.5-122B-A10B: 58%Normalized gap 17.0GPT-5.4 source Qwen3.5-122B-A10B source
GPQA
Knowledge
GPT-5.4: 92.8%Qwen3.5-122B-A10B: 86.6%Normalized gap 6.2GPT-5.4 source Qwen3.5-122B-A10B source
CharXiv
Multimodal
GPT-5.4: 82.8%Qwen3.5-122B-A10B: 77.2%Normalized gap 5.6GPT-5.4 source Qwen3.5-122B-A10B 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

GPT-5.4: $0.01; Fits in one request
Qwen3.5-122B-A10B: Self-hosted; infrastructure cost varies; Fits in one request

Qwen3.5-122B-A10B has no comparable published API token rate.

Repository review

50K fresh input + 3K output tokens

GPT-5.4: $0.17; Fits in one request
Qwen3.5-122B-A10B: Self-hosted; infrastructure cost varies; Fits in one request

Qwen3.5-122B-A10B has no comparable published API token rate.

Cache-heavy agent loop

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

GPT-5.4: $0.25; Fits in one request
Qwen3.5-122B-A10B: Self-hosted; infrastructure cost varies; Fits in one request; Cached-input rate unavailable

Qwen3.5-122B-A10B 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.

SpecificationGPT-5.4Qwen3.5-122B-A10B

Context window

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

GPT-5.4

1.05M

OpenAI pricing

Qwen3.5-122B-A10B

262K

API model ID

GPT-5.4

gpt-5.4

OpenAI pricing

Qwen3.5-122B-A10B

Not sourced

Cached-input rate

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

GPT-5.4

$0.25 per 1M cached input tokens

OpenAI pricing

Qwen3.5-122B-A10B

No comparable hosted API rate

Documented inputs

GPT-5.4

Not sourced

Qwen3.5-122B-A10B

Not sourced

Documented outputs

GPT-5.4

Not sourced

Qwen3.5-122B-A10B

Not sourced

Provider availability

GPT-5.4

Not sourced

Qwen3.5-122B-A10B

Not sourced

Reasoning profile

GPT-5.4

Reasoning

Qwen3.5-122B-A10B

Reasoning

Weight access

GPT-5.4

Proprietary

Qwen3.5-122B-A10B

Open Weight

License

GPT-5.4

Proprietary

Qwen3.5-122B-A10B

Open Weight

Release date

GPT-5.4

2026-03-05

Qwen3.5-122B-A10B

2026-03-04

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: GPT-5.4 has the higher public score estimate, 73.16 versus 59.47, but the 90% score intervals overlap.
Workload cost: A complete comparable API-rate estimate is not available for both models.
Context tradeoff: GPT-5.4 has the larger documented window (1.05M).

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 evidence48 rows

Agentic

Terminal-Bench 2.0
GPT-5.475.1%
Source
Qwen3.5-122B-A10B49.4%
Source
GPT-5.4 leads this result
CyberGym
GPT-5.479.0%
Source
Qwen3.5-122B-A10B—
Not directly comparable
BrowseComp
GPT-5.482.7%
Source
Qwen3.5-122B-A10B63.8%
Source
GPT-5.4 leads this result
OSWorld-Verified
GPT-5.475%
Source
Qwen3.5-122B-A10B58%
Source
GPT-5.4 leads this result
MCP Atlas
GPT-5.470.6%
Source
Qwen3.5-122B-A10B—
Not directly comparable
Toolathlon
GPT-5.454.6%
Source
Qwen3.5-122B-A10B—
Not directly comparable
τ²-bench results
GPT-5.498.9%
Source
Qwen3.5-122B-A10B—
Not directly comparable
Claw-Eval
GPT-5.460.3%
Source
Qwen3.5-122B-A10B—
Not directly comparable
DeepSearchQA
GPT-5.473.6%
Source
Qwen3.5-122B-A10B—
Not directly comparable
Gert Labs
GPT-5.464.89%
Source
Qwen3.5-122B-A10B—
Not directly comparable
ResearchClawBench
GPT-5.415.3%
Source
Qwen3.5-122B-A10B—
Not directly comparable
JobBench
GPT-5.438.9%
Source
Qwen3.5-122B-A10B—
Not directly comparable
ExploitGym
GPT-5.46.0%
Source
Qwen3.5-122B-A10B—
Not directly comparable

Coding

LiveCodeBench Pro
GPT-5.487.5%
Source
Qwen3.5-122B-A10B—
Not directly comparable
SWE-bench Pro
GPT-5.457.7%
Source
Qwen3.5-122B-A10B—
Not directly comparable
React Native Evals
GPT-5.485.3%
Source
Qwen3.5-122B-A10B—
Not directly comparable
Vibe Code Bench
GPT-5.467.42%
Source
Qwen3.5-122B-A10B—
Not directly comparable
EEBench
GPT-5.438.5%
Source
Qwen3.5-122B-A10B—
Not directly comparable
SWE-bench Verified
GPT-5.4—
Qwen3.5-122B-A10B72%
Source
Not directly comparable

Reasoning

ARC-AGI-2
GPT-5.474.0%
Source
Qwen3.5-122B-A10B—
Not directly comparable
ARC-AGI-3
GPT-5.40.2%
Source
Qwen3.5-122B-A10B—
Not directly comparable
LongBench v2
GPT-5.4—
Qwen3.5-122B-A10B60.2%
Source
Not directly comparable

Knowledge

GPQA
GPT-5.492.8%
Source
Qwen3.5-122B-A10B86.6%
Source
GPT-5.4 leads this result
HLE
GPT-5.452.1%
Source
Qwen3.5-122B-A10B—
Not directly comparable
HLE w/o tools
GPT-5.439.8%
Source
Qwen3.5-122B-A10B—
Not directly comparable
GPQA-D
GPT-5.492.8%
Source
Qwen3.5-122B-A10B—
Not directly comparable
HealthBench Hard
GPT-5.440.1%
Source
Qwen3.5-122B-A10B—
Not directly comparable
MedXpertQA (Text)
GPT-5.459.6%
Source
Qwen3.5-122B-A10B—
Not directly comparable
HealthBench Professional
GPT-5.448.1%
Source
Qwen3.5-122B-A10B—
Not directly comparable
MMLU-Pro
GPT-5.4—
Qwen3.5-122B-A10B86.7%
Source
Not directly comparable
SuperGPQA
GPT-5.4—
Qwen3.5-122B-A10B67.1%
Source
Not directly comparable

Math

FrontierMath v2 (Tiers 1-3)
GPT-5.447.600%
Source
Qwen3.5-122B-A10B—
Not directly comparable
FrontierMath v2 (Tier 4)
GPT-5.427.100%
Source
Qwen3.5-122B-A10B—
Not directly comparable

Multilingual

MMLU-ProX
GPT-5.4—
Qwen3.5-122B-A10B82.2%
Source
Not directly comparable

Multimodal

MMMU-Pro
GPT-5.481.2%
Source
Qwen3.5-122B-A10B—
Not directly comparable
OfficeQA Pro
GPT-5.453.2%
Source
Qwen3.5-122B-A10B—
Not directly comparable
MMMU-Pro w/ Python
GPT-5.482.1%
Source
Qwen3.5-122B-A10B—
Not directly comparable
CharXiv
GPT-5.482.8%
Source
Qwen3.5-122B-A10B77.2%
Source
GPT-5.4 leads this result
ERQA
GPT-5.465.4%
Source
Qwen3.5-122B-A10B—
Not directly comparable
SimpleVQA
GPT-5.461.1%
Source
Qwen3.5-122B-A10B—
Not directly comparable
ScreenSpot Pro
GPT-5.485.4%
Source
Qwen3.5-122B-A10B—
Not directly comparable
ZeroBench
GPT-5.441.0%
Source
Qwen3.5-122B-A10B—
Not directly comparable
MedXpertQA (MM)
GPT-5.477.1%
Source
Qwen3.5-122B-A10B—
Not directly comparable
MMMU
GPT-5.4—
Qwen3.5-122B-A10B83.9%
Source
Not directly comparable
MMVU
GPT-5.4—
Qwen3.5-122B-A10B74.7%
Source
Not directly comparable
MathVision
GPT-5.4—
Qwen3.5-122B-A10B86.2%
Source
Not directly comparable
V*
GPT-5.4—
Qwen3.5-122B-A10B93.2%
Source
Not directly comparable

Instruction following

IFEval
GPT-5.4—
Qwen3.5-122B-A10B93.4%
Source
Not directly comparable

Frequently asked questions

Which is better, GPT-5.4 or Qwen3.5-122B-A10B?

GPT-5.4 has the higher public score estimate, 73.16 versus 59.47, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.

Which is better for coding, GPT-5.4 or Qwen3.5-122B-A10B?

The published evidence does not provide a shared weighted coding basis for both models, so BenchLM does not name a coding winner.

Which is better for agentic tasks, GPT-5.4 or Qwen3.5-122B-A10B?

GPT-5.4 leads the like-for-like agentic tasks comparison across 3 shared weighted benchmark rows.

Which costs less, GPT-5.4 or Qwen3.5-122B-A10B?

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, GPT-5.4 or Qwen3.5-122B-A10B?

GPT-5.4 has the larger documented context window: 1.05M, compared with 262K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated August 13, 2026

Watch GPT-5.4 vs Qwen3.5-122B-A10B

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

Read a sample issue

Join 2,000+ readers.

GPT-5.4 vs Qwen3.5-122B-A10B

Which one for your work

Agentic work

Long documents

Coding work

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

What is actually comparable

Category results, on a stated basis

Agentic

Knowledge

Multimodal

Coding

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, GPT-5.4 or Qwen3.5-122B-A10B?

Which is better for coding, GPT-5.4 or Qwen3.5-122B-A10B?

Which is better for agentic tasks, GPT-5.4 or Qwen3.5-122B-A10B?

Which costs less, GPT-5.4 or Qwen3.5-122B-A10B?

Which has the larger context window, GPT-5.4 or Qwen3.5-122B-A10B?

Related comparisons

Watch GPT-5.4 vs Qwen3.5-122B-A10B