Agentic work
Tool use, computer use, and multi-step task completion
GPT-5.2
GPT-5.2 leads on the public agentic lane, 42.7 to 35.2, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
GPT-5.2 vs Qwen3.7 Plus
Updated September 29, 2026. Rank says GPT-5.2 is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Both of these models will change. Get the price, version and retirement notices for the pair, each with its source. Follow model changes
Share or export
Decision reading
GPT-5.2 has the higher public score estimate, 61.18 versus 55.73, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 7 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 7
- GPT-5.2 only
- 8
- Qwen3.7 Plus only
- 45
- Like-for-like categories
- 2 / 8
Supported: GPT-5.2 and Qwen3.7 PlusHow the comparison works
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
Qwen3.7 Plus
Qwen3.7 Plus 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
Qwen3.7 Plus is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limitedChat 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-ratesCache-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-fallbackRepository 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
Which one for a specific job
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
39.7GPT-5.243.1Qwen3.7 Plus
Directional only · BenchAlign v5.7
Qwen3.7 Plus scores higher, but at least one score rests on Estimated evidence or a different benchmark set. Directional only, no winner.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Same basis rules as the category table below
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
8 GPT-5.2 only
7 Shared
45 Qwen3.7 Plus only
3 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
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-VerifiedAgentic
Normalized gap 26.0GPT-5.2:47.3%Qwen3.7 Plus:73.3%CharXivMultimodal
Normalized gap 3.8GPT-5.2:82.1%Qwen3.7 Plus:85.9%SWE-bench VerifiedCoding
Normalized gap 2.3GPT-5.2:80%Qwen3.7 Plus:77.7%GPQAKnowledge
Normalized gap 2.1GPT-5.2:92.4%Qwen3.7 Plus:90.3%SWE-bench ProCoding
Normalized gap 2.0GPT-5.2:55.6%Qwen3.7 Plus:57.6%
Bars run 0–100 on each benchmark’s normalized display scale
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.7 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Agentic
Like-for-like- GPT-5.2
- 42.7
- Supported · #53/117
- Qwen3.7 Plus
- 35.2
- Supported · #66/117
- Basis
- BenchAlign v5.7 lane · 4 vs 11 public rows
- Reading
- GPT-5.2 leads · intervals overlap
Multimodal
Like-for-like- GPT-5.2
- 67.3
- #23/50
- Qwen3.7 Plus
- 73.5
- #19/50
- Basis
- Provisional lane · 2 vs 2 weighted rows
- Reading
- Qwen3.7 Plus leads
Coding
Directional only- GPT-5.2
- 39.7
- Supported · #62/143
- Qwen3.7 Plus
- 43.1
- Estimated · #55/143
- Basis
- BenchAlign v5.7 lane · 3 vs 7 public rows
- Reading
- Directional only
Knowledge
Directional only- GPT-5.2
- 57.6
- Supported · #44/169
- Qwen3.7 Plus
- 52.3
- Estimated · #55/169
- Basis
- BenchAlign v5.7 lane · 1 vs 7 public rows
- Reading
- Directional only
Instruction following
Directional only- GPT-5.2
- 91.2
- #15/124
- Qwen3.7 Plus
- 89.2
- #18/124
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Directional only
Reasoning
Not comparable- GPT-5.2
- 60.7
- Unranked · 3 rankable rows
- Qwen3.7 Plus
- 75.2
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 1 vs 1 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- GPT-5.2
- Not ranked
- Qwen3.7 Plus
- 78.9
- #3/12
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Math
Not comparable- GPT-5.2
- 57.4
- Unranked · 2 rankable rows
- Qwen3.7 Plus
- 78.2
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 2 vs 1 weighted rows
- Reading
- Not comparable
| Category | GPT-5.2 | Qwen3.7 Plus | Basis | Reading |
|---|---|---|---|---|
| Agentic | 42.7Supported · #53/117 | 35.2Supported · #66/117 | Like-for-likeBenchAlign v5.7 lane · 4 vs 11 public rows | GPT-5.2 leads · intervals overlap |
| Multimodal | 67.3#23/50 | 73.5#19/50 | Like-for-likeProvisional lane · 2 vs 2 weighted rows | Qwen3.7 Plus leads |
| Coding | 39.7Supported · #62/143 | 43.1Estimated · #55/143 | Directional onlyBenchAlign v5.7 lane · 3 vs 7 public rows | Directional only |
| Knowledge | 57.6Supported · #44/169 | 52.3Estimated · #55/169 | Directional onlyBenchAlign v5.7 lane · 1 vs 7 public rows | Directional only |
| Instruction following | 91.2#15/124 | 89.2#18/124 | Directional onlyProvisional lane · 0 vs 1 weighted rows | Directional only |
| Reasoning | 60.7Unranked · 3 rankable rows | 75.2Unranked · 3 rankable rows | Not comparableProvisional lane · 1 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | 78.9#3/12 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Math | 57.4Unranked · 2 rankable rows | 78.2Unranked · 3 rankable rows | Not comparableProvisional lane · 2 vs 1 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Supported evidence per lane · bars run 0–100Methodology
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-5.2
- $0.00875
- Fits in one request
- Qwen3.7 Plus
- API rate not published
- Fits in one request
Qwen3.7 Plus has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- GPT-5.2
- $0.1295
- Fits in one request
- Qwen3.7 Plus
- API rate not published
- Fits in one request
Qwen3.7 Plus has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- GPT-5.2
- $0.525
- Fits in one request
- Cached input priced at the published list-input rate
- Qwen3.7 Plus
- API rate not published
- Fits in one request
- Cached-input rate unavailable
GPT-5.2 has no published cached-input rate, so cached tokens use its listed input rate. Qwen3.7 Plus has no comparable published API token rate.
Cached input falls back to the list input rate only where a cached rate is unpublished
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationGPT-5.2Qwen3.7 Plus
Context window
Maximum documented context; output-token limits may be lower.
GPT-5.2
400K
Qwen3.7 Plus
1M
API model ID
GPT-5.2
Not sourced
Qwen3.7 Plus
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-5.2
Not published
Qwen3.7 Plus
No comparable hosted API rate
Documented inputs
GPT-5.2
Not sourced
Qwen3.7 Plus
Not sourced
Documented outputs
GPT-5.2
Not sourced
Qwen3.7 Plus
Not sourced
Provider availability
GPT-5.2
Not sourced
Qwen3.7 Plus
Not sourced
Reasoning profile
GPT-5.2
Reasoning
Qwen3.7 Plus
Reasoning
Weight access
GPT-5.2
Proprietary
Qwen3.7 Plus
Proprietary
License
GPT-5.2
Proprietary
Qwen3.7 Plus
Proprietary
Release date
GPT-5.2
2025-12-11
Qwen3.7 Plus
2026-06-03
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-5.2 has the higher public score estimate, 61.18 versus 55.73, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Qwen3.7 Plus has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Questions
Which is better, GPT-5.2 or Qwen3.7 Plus?
GPT-5.2 has the higher public score estimate, 61.18 versus 55.73, 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-5.2 or Qwen3.7 Plus?
Qwen3.7 Plus scores higher for coding on the public lane, 43.1 to 39.7. Qwen3.7 Plus is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, GPT-5.2 or Qwen3.7 Plus?
GPT-5.2 leads the public agentic tasks lane, 42.7 to 35.2, with Supported evidence for both models, although the 90% intervals overlap.
Which costs less, GPT-5.2 or Qwen3.7 Plus?
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-5.2 or Qwen3.7 Plus?
Qwen3.7 Plus has the larger documented context window: 1M, compared with 400K.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence60 rows
Agentic
BrowseComp
GPT-5.265.8%SourceQwen3.7 Plus—Not directly comparable
OSWorld-Verified
GPT-5.247.3%SourceQwen3.7 Plus73.3%SourceQwen3.7 Plus leads this result
Gert Labs
GPT-5.246.54%SourceQwen3.7 Plus—Not directly comparable
JobBench
GPT-5.234.3%SourceQwen3.7 Plus—Not directly comparable
Terminal-Bench 2.0
GPT-5.2—Qwen3.7 Plus70.3%SourceNot directly comparable
QwenClawBench
GPT-5.2—Qwen3.7 Plus61.8%SourceNot directly comparable
Claw-Eval
GPT-5.2—Qwen3.7 Plus62.7%SourceNot directly comparable
BFCL v4
GPT-5.2—Qwen3.7 Plus72.9%SourceNot directly comparable
MCP Atlas
GPT-5.2—Qwen3.7 Plus73.2%SourceNot directly comparable
VITA-Bench
GPT-5.2—Qwen3.7 Plus45.6%SourceNot directly comparable
DeepPlanning
GPT-5.2—Qwen3.7 Plus62.3%SourceNot directly comparable
AndroidWorld
GPT-5.2—Qwen3.7 Plus81.0%SourceNot directly comparable
OSWorld 2.0
GPT-5.2—Qwen3.7 Plus2.8%SourceNot directly comparable
Terminal-Bench 2.1 (Vals)
GPT-5.2—Qwen3.7 Plus52.8%SourceNot directly comparable
Coding
SWE-bench Verified
GPT-5.280%SourceQwen3.7 Plus77.7%SourceGPT-5.2 leads this result
SWE-bench Pro
GPT-5.255.6%SourceQwen3.7 Plus57.6%SourceQwen3.7 Plus leads this result
Vibe Code Bench
GPT-5.253.50%SourceQwen3.7 Plus—Not directly comparable
Terminal-Bench 2.0
GPT-5.2—Qwen3.7 Plus70.3%SourceNot directly comparable
SWE Multilingual
GPT-5.2—Qwen3.7 Plus75.8%SourceNot directly comparable
NL2Repo
GPT-5.2—Qwen3.7 Plus41.1%SourceNot directly comparable
SciCode
GPT-5.2—Qwen3.7 Plus51.3%SourceNot directly comparable
LiveCodeBench
GPT-5.2—Qwen3.7 Plus89.6%SourceNot directly comparable
Reasoning
Multimodal
MMMU-Pro
GPT-5.279.5%SourceQwen3.7 Plus79%SourceGPT-5.2 leads this result
MathVision
GPT-5.283.0%SourceQwen3.7 Plus90.3%SourceQwen3.7 Plus leads this result
CharXiv
GPT-5.282.1%SourceQwen3.7 Plus85.9%SourceQwen3.7 Plus leads this result
V*
GPT-5.275.9%SourceQwen3.7 Plus—Not directly comparable
ERQA
GPT-5.2—Qwen3.7 Plus69.8%SourceNot directly comparable
MedXpertQA (MM)
GPT-5.2—Qwen3.7 Plus71.0%SourceNot directly comparable
ScreenSpot Pro
GPT-5.2—Qwen3.7 Plus79.0%SourceNot directly comparable
SimpleVQA
GPT-5.2—Qwen3.7 Plus81.7%SourceNot directly comparable
MMSearch-Plus
GPT-5.2—Qwen3.7 Plus41.4%SourceNot directly comparable
RealWorldQA
GPT-5.2—Qwen3.7 Plus86.9%SourceNot directly comparable
OmniDocBench 1.5
GPT-5.2—Qwen3.7 Plus91.4%SourceNot directly comparable
OCRBench V2
GPT-5.2—Qwen3.7 Plus70.7%SourceNot directly comparable
ODINW13
GPT-5.2—Qwen3.7 Plus51.1%SourceNot directly comparable
Video-MME (with subtitle)
GPT-5.2—Qwen3.7 Plus88.0%SourceNot directly comparable
VideoMMMU
GPT-5.2—Qwen3.7 Plus85.4%SourceNot directly comparable
MLVU (M-Avg)
GPT-5.2—Qwen3.7 Plus87.4%SourceNot directly comparable
Knowledge
GPQA
GPT-5.292.4%SourceQwen3.7 Plus90.3%SourceGPT-5.2 leads this result
GPQA-D
GPT-5.2—Qwen3.7 Plus90.3%SourceNot directly comparable
HLE
GPT-5.2—Qwen3.7 Plus34.7%SourceNot directly comparable
MMLU-Pro
GPT-5.2—Qwen3.7 Plus88.5%SourceNot directly comparable
MMLU-Redux
GPT-5.2—Qwen3.7 Plus94.5%SourceNot directly comparable
SuperGPQA
GPT-5.2—Qwen3.7 Plus71.4%SourceNot directly comparable
MMMLU
GPT-5.2—Qwen3.7 Plus89.0%SourceNot directly comparable
Multilingual
MMLU-ProX
GPT-5.2—Qwen3.7 Plus85.4%SourceNot directly comparable
NOVA-63
GPT-5.2—Qwen3.7 Plus58.8%SourceNot directly comparable
INCLUDE
GPT-5.2—Qwen3.7 Plus83.0%SourceNot directly comparable
MAXIFE
GPT-5.2—Qwen3.7 Plus88.8%SourceNot directly comparable
PolyMath
GPT-5.2—Qwen3.7 Plus84.0%SourceNot directly comparable
Instruction following
Math
FrontierMath v2 (Tiers 1-3)
GPT-5.240.700%SourceQwen3.7 Plus—Not directly comparable
FrontierMath v2 (Tier 4)
GPT-5.218.800%SourceQwen3.7 Plus—Not directly comparable
HMMT Feb 2026
GPT-5.2—Qwen3.7 Plus92.9%SourceNot directly comparable
IMOAnswerBench
GPT-5.2—Qwen3.7 Plus86.0%SourceNot directly comparable
Apex
GPT-5.2—Qwen3.7 Plus22.7%SourceNot directly comparable
60 public results · 7 shared
Watch GPT-5.2 vs Qwen3.7 Plus
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.
Last updated September 29, 2026