Back

OpenAI

GPT-5.5

ReasoningClosed weights
Intelligence
37.0
Rank 84 of 744 measured models
Coding
71.6
Rank 45 of 298 measured models
Output speed
82.7 t/s
Rank 221 of 403 measured models
Price/1M
$11.25
Rank 630 of 659 measured models
Context window
922K
Rank 330 of 1199 measured models
Cost per task
$1.54
Rank 152 of 193 measured models

GPT-5.5 (High) scores 37.0 on the Artificial Analysis Intelligence Index, ranking 84th of 744 measured models. The closest models on this index are MiMo-V2.6-Flash (37.9), GPT-5.6 Luna (37.3), and DeepSeek V4 Pro 0813 (36.0). At $11.25 per million tokens (3:1 input/output blended price), it is more expensive than MiMo-V2.6-Flash ($0.175), GPT-5.6 Luna ($0.45), and DeepSeek V4 Pro 0813 ($1.98). With an output speed of 82.7 t/s, it is faster than MiMo-V2.6-Flash (53.8 t/s), slower than GPT-5.6 Luna (115 t/s) and DeepSeek V4 Pro 0813 (92 t/s).

Artificial Analysis Indices
High
Intelligence37.0
Rank 84 of 744 measured models
Coding71.6
Rank 45 of 298 measured models
Agentic34.3
Rank 60 of 189 measured models
Reasoning levels
Each point is a reasoning level of this model: more reasoning often raises the score but costs more. Select a point to open that level.
Intelligence Indexmost efficient ↗Cost per task
Capability indexes
Performance in specific industries, measured by Artificial Analysis
Finance & Accounting40.1
Rank 76 of 218 measured models
Strategy & Ops43.5
Rank 71 of 218 measured models
Legal43.5
Rank 67 of 218 measured models
Engineering38.6
Rank 73 of 217 measured models
Economics46.7
Rank 73 of 225 measured models
Standard Benchmarks
GPQA Diamond93.2%
Rank 22 of 619 measured models
HLE45.0%
Rank 49 of 664 measured models
SciCode56.1%
Rank 40 of 231 measured models
IFBench71.6%
Rank 53 of 450 measured models
AA-LCR84.3%
Rank 8 of 575 measured models
TerminalBench Hard59.8%
Rank 9 of 432 measured models
Terminal-Bench 2.179.4%
Rank 50 of 244 measured models
Terminal-Bench v49.1%
Rank 89 of 218 measured models
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.

Preferences and analytics

Rybbit: visits, performance, clicks, copies, forms and session replay (masked inputs), with your permission. Refusing does not limit the site.

Privacy

Allow navigation analytics?

With your permission, Rybbit measures visits, performance and interactions (clicks, copies, forms), including session replay, to improve BenchSift. Input values are masked in replays. Refusing does not limit the site.

Your choice is stored for 180 days. Change it at any time using “Privacy preferences” in the footer. Withdrawal reloads the page.

No choice stored: Rybbit is disabled.

Privacy