Agentic
Directional only- GPT-5.3 Codex
- 71.4
- Qwen3.8 Max
- 86.1
- Weighted basis
- 2 vs 1 rows
- Reading
- Directional only
Model comparison
GPT-5.3 Codex vs Qwen3.8 Max
Updated August 3, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
GPT-5.3 Codex has the higher public score estimate, 65.7 versus 65.4, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
3 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.
Long documents
Prompts that approach the documented context limit
Qwen3.8 Max
Qwen3.8 Max 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
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
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: rate-fallback
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.
5 GPT-5.3 Codex only
3 Shared
49 Qwen3.8 Max only
- Shared results
- 3
- GPT-5.3 Codex only
- 5
- Qwen3.8 Max only
- 49
- Like-for-like categories
- 0 / 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.
Coding
Directional only- GPT-5.3 Codex
- 67.2
- Qwen3.8 Max
- 67.7
- Weighted basis
- 3 vs 1 rows
- Reading
- Directional only
Reasoning
Not comparable- GPT-5.3 Codex
- Not measured
- Qwen3.8 Max
- 78.3
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Knowledge
Not comparable- GPT-5.3 Codex
- Not measured
- Qwen3.8 Max
- 50.2
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Math
Not comparable- GPT-5.3 Codex
- Not measured
- Qwen3.8 Max
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- GPT-5.3 Codex
- Not measured
- Qwen3.8 Max
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- GPT-5.3 Codex
- Not measured
- Qwen3.8 Max
- 86.3
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- GPT-5.3 Codex
- Not measured
- Qwen3.8 Max
- 82.8
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
| Category | GPT-5.3 Codex | Qwen3.8 Max | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | 71.4 | 86.1 | Directional only2 vs 1 rows | Directional only |
| Coding | 67.2 | 67.7 | Directional only3 vs 1 rows | Directional only |
| Reasoning | Not measured | 78.3 | Not comparable0 vs 2 rows | Not comparable |
| Knowledge | Not measured | 50.2 | Not comparable0 vs 2 rows | Not comparable |
| Math | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 86.3 | Not comparable0 vs 2 rows | Not comparable |
| Instruction following | Not measured | 82.8 | 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.
OSWorld-Verified
Agentic
SWE-bench Pro
Coding
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-5.3 Codex
- $0.00875
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
Qwen3.8 Max has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- GPT-5.3 Codex
- $0.1295
- Fits in one request
- Qwen3.8 Max
- API rate not published
- Fits in one request
Qwen3.8 Max has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- GPT-5.3 Codex
- $0.525
- Fits in one request
- Cached input priced at the published list-input rate
- Qwen3.8 Max
- API rate not published
- Fits in one request
- Cached-input rate unavailable
GPT-5.3 Codex has no published cached-input rate, so cached tokens use its listed input rate. Qwen3.8 Max 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-5.3 CodexQwen3.8 Max
Context window
Maximum documented context; output-token limits may be lower.
GPT-5.3 Codex
Qwen3.8 Max
API model ID
GPT-5.3 Codex
gpt-5.3-codex
OpenAI GPT-5.3 Codex model documentationQwen3.8 Max
qwen3.8-max
Alibaba Cloud Model Studio pricingCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-5.3 Codex
Not published
Qwen3.8 Max
No comparable hosted API rate
Alibaba Cloud Model Studio pricingDocumented inputs
GPT-5.3 Codex
text, image
OpenAI model catalogQwen3.8 Max
Not sourced
Documented outputs
GPT-5.3 Codex
Qwen3.8 Max
Not sourced
Provider availability
GPT-5.3 Codex
Generally Available · OpenAI Responses API
OpenAI model catalogQwen3.8 Max
Not sourced
Reasoning profile
GPT-5.3 Codex
Reasoning
Qwen3.8 Max
Reasoning
Weight access
GPT-5.3 Codex
Proprietary
Qwen3.8 Max
Proprietary
License
GPT-5.3 Codex
Proprietary
Qwen3.8 Max
Proprietary
Release date
GPT-5.3 Codex
2026-02-05
Qwen3.8 Max
2026-08-03
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-5.3 Codex has the higher public score estimate, 65.7 versus 65.4, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Qwen3.8 Max has the larger documented window (1M).
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 evidence57 rows
Agentic
Terminal-Bench 2.0
GPT-5.3 Codex77.3%SourceQwen3.8 Max—Not directly comparable
OSWorld-Verified
GPT-5.3 Codex64.7%SourceQwen3.8 Max86.1%SourceQwen3.8 Max leads this result
Gert Labs
GPT-5.3 Codex57.47%SourceQwen3.8 Max—Not directly comparable
JobBench
GPT-5.3 Codex33.7%SourceQwen3.8 Max53.4%SourceQwen3.8 Max leads this result
Terminal-Bench 2.1
GPT-5.3 Codex—Qwen3.8 Max86.6%SourceNot directly comparable
CoWorkBench
GPT-5.3 Codex—Qwen3.8 Max74.8%SourceNot directly comparable
skillsBench
GPT-5.3 Codex—Qwen3.8 Max70.2%SourceNot directly comparable
Agents' Last Exam
GPT-5.3 Codex—Qwen3.8 Max52.4%SourceNot directly comparable
AutomationBench
GPT-5.3 Codex—Qwen3.8 Max27.3%SourceNot directly comparable
Toolathlon-Verified
GPT-5.3 Codex—Qwen3.8 Max72.5%SourceNot directly comparable
WideResearch
GPT-5.3 Codex—Qwen3.8 Max81.9%SourceNot directly comparable
HLE w/ tools
GPT-5.3 Codex—Qwen3.8 Max56.2%SourceNot directly comparable
OSWorld 2.0
GPT-5.3 Codex—Qwen3.8 Max19.4%SourceNot directly comparable
WebArena-Verified
GPT-5.3 Codex—Qwen3.8 Max66.8%SourceNot directly comparable
AndroidWorld
GPT-5.3 Codex—Qwen3.8 Max85.3%SourceNot directly comparable
MobileWorld
GPT-5.3 Codex—Qwen3.8 Max77.8%SourceNot directly comparable
Coding
SWE-bench Verified
GPT-5.3 Codex85%SourceQwen3.8 Max—Not directly comparable
SWE-bench Pro
GPT-5.3 Codex56.8%SourceQwen3.8 Max67.7%SourceQwen3.8 Max leads this result
SWE-Rebench
GPT-5.3 Codex58.2%SourceQwen3.8 Max—Not directly comparable
Vibe Code Bench
GPT-5.3 Codex61.77%SourceQwen3.8 Max—Not directly comparable
Terminal-Bench 2.1
GPT-5.3 Codex—Qwen3.8 Max86.6%SourceNot directly comparable
deepSwe
GPT-5.3 Codex—Qwen3.8 Max56.6%SourceNot directly comparable
NL2Repo
GPT-5.3 Codex—Qwen3.8 Max55.9%SourceNot directly comparable
FrontierSWE
GPT-5.3 Codex—Qwen3.8 Max73.5%SourceNot directly comparable
MLS-Bench Lite
GPT-5.3 Codex—Qwen3.8 Max41.0%SourceNot directly comparable
PaperBench
GPT-5.3 Codex—Qwen3.8 Max93.0%SourceNot directly comparable
Reasoning
Knowledge
Multimodal
MMMU-Pro
GPT-5.3 Codex—Qwen3.8 Max82.3%SourceNot directly comparable
MathVision
GPT-5.3 Codex—Qwen3.8 Max95.2%SourceNot directly comparable
MathVision w/ Python
GPT-5.3 Codex—Qwen3.8 Max97.7%SourceNot directly comparable
BabyVision
GPT-5.3 Codex—Qwen3.8 Max82.0%SourceNot directly comparable
BabyVision w/ Python
GPT-5.3 Codex—Qwen3.8 Max91.3%SourceNot directly comparable
ZeroBench
GPT-5.3 Codex—Qwen3.8 Max24.0%SourceNot directly comparable
ZeroBench w/ Python
GPT-5.3 Codex—Qwen3.8 Max49.0%SourceNot directly comparable
MedXpertQA (MM)
GPT-5.3 Codex—Qwen3.8 Max80.4%SourceNot directly comparable
ScreenSpot Pro
GPT-5.3 Codex—Qwen3.8 Max84.5%SourceNot directly comparable
Vision2Web
GPT-5.3 Codex—Qwen3.8 Max69.0%SourceNot directly comparable
CharXiv w/o tools
GPT-5.3 Codex—Qwen3.8 Max88.4%SourceNot directly comparable
CharXiv
GPT-5.3 Codex—Qwen3.8 Max93.5%SourceNot directly comparable
OmniDocBench 1.5
GPT-5.3 Codex—Qwen3.8 Max92.1%SourceNot directly comparable
OCRBench V2
GPT-5.3 Codex—Qwen3.8 Max74.2%SourceNot directly comparable
CC-OCR
GPT-5.3 Codex—Qwen3.8 Max79.6%SourceNot directly comparable
RealWorldQA
GPT-5.3 Codex—Qwen3.8 Max88.0%SourceNot directly comparable
ERQA
GPT-5.3 Codex—Qwen3.8 Max77.8%SourceNot directly comparable
SimpleVQA
GPT-5.3 Codex—Qwen3.8 Max75.0%SourceNot directly comparable
PerceptionBench
GPT-5.3 Codex—Qwen3.8 Max63.5%SourceNot directly comparable
Video-MME (with subtitle)
GPT-5.3 Codex—Qwen3.8 Max90.4%SourceNot directly comparable
VideoMMMU
GPT-5.3 Codex—Qwen3.8 Max88.7%SourceNot directly comparable
MMVU
GPT-5.3 Codex—Qwen3.8 Max82.4%SourceNot directly comparable
MLVU (M-Avg)
GPT-5.3 Codex—Qwen3.8 Max90.8%SourceNot directly comparable
LVBench
GPT-5.3 Codex—Qwen3.8 Max81.8%SourceNot directly comparable
Instruction following
IFBench
GPT-5.3 Codex—Qwen3.8 Max82.8%SourceNot directly comparable
Frequently asked questions
Which is better, GPT-5.3 Codex or Qwen3.8 Max?
GPT-5.3 Codex has the higher public score estimate, 65.7 versus 65.4, 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-5.3 Codex or Qwen3.8 Max?
The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, GPT-5.3 Codex or Qwen3.8 Max?
The current agentic tasks averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which costs less, GPT-5.3 Codex or Qwen3.8 Max?
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-5.3 Codex or Qwen3.8 Max?
Qwen3.8 Max has the larger documented context window: 1M, compared with 400K.
Related comparisons
Last updated August 3, 2026
Watch GPT-5.3 Codex vs Qwen3.8 Max
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.