← All models
Compare

DeepSeek-V4-Pro vs GPT-5.5

Add, remove, or swap models to compare them side by side.

DeepSeek-V4-ProGPT-5.5
DeepSeek-V4-ProDeepSeek · 3 in the newsGPT-5.5OpenAI · 17 in the news
Scores
Intelligence (ECI)150159
Coding7691
Math9197
Reasoning & Knowledge6892
Agentic & Tools9287
Specifications
DeveloperDeepSeekOpenAI
FamilyDeepSeekGPT
ReleasedApr 24, 2026Apr 23, 2026
Parameters1.6T
AvailabilityOpen weights (unrestricted)API access
Context window1M1.1M
Price — $/M input$0.43$5.00
Price — $/M output$0.87$30.00
Inputstextfile, image, text
Outputstexttext
Benchmarks
AIME 2024/202597%100%
GPQA Diamond90%94%
Humanity's Last Exam36%44%
SciCode50%56%
SimpleBench61%69%
SimpleQA Verified57%63%
SWE-bench Verified78%81%
WebDev Arena14641505
WeirdML49%85%
τ²-bench96%94%
APEX38%
ARC-AGI95%
ARC-AGI-285%
FrontierMath52%
FrontierMath Tier 435%
GSO (code optimization)40%
Terminal-Bench85%

Use-case scores are 0–100 percentile composites across each area’s benchmarks, ranked against every model from the past year. Highlighted cells lead each row. Open a model for the full picture.

Benchmarks & model data from Epoch AI (CC BY); pricing & specs from OpenRouter. ECI = Epoch Capabilities Index.