Back

OpenAI

GPT-5.6 Terra

Closed weights
Intelligence
20.8
Rank 232 of 744 measured models
Coding
52.3
Rank 114 of 298 measured models
Output speed
90.1 t/s
Rank 197 of 403 measured models
Price/1M
$4.50
Rank 563 of 659 measured models
Context window
1M
Rank 84 of 1199 measured models
Cost per task
$0.14
Rank 63 of 193 measured models

GPT-5.6 Terra (Non-reasoning) scores 20.8 on the Artificial Analysis Intelligence Index, ranking 232nd of 744 measured models. The closest models on this index are MiniMax-M2.1 (20.9), GPT-5.4 nano (20.7), and GPT-5 Mini (batch) (20.6). At $4.50 per million tokens (3:1 input/output blended price), it is more expensive than MiniMax-M2.1 ($0.525) and GPT-5.4 nano ($0.463). With an output speed of 90.1 t/s, it is faster than MiniMax-M2.1 (62.2 t/s), slower than GPT-5.4 nano (166 t/s).

Artificial Analysis Indices
Non-reasoning
Intelligence20.8
Rank 232 of 744 measured models
Coding52.3
Rank 114 of 298 measured models
Agentic23.5
Rank 89 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 & Accounting21.0
Rank 154 of 218 measured models
Strategy & Ops23.0
Rank 135 of 218 measured models
Legal22.2
Rank 155 of 218 measured models
Engineering21.1
Rank 157 of 217 measured models
Economics23.9
Rank 170 of 225 measured models
Standard Benchmarks
GPQA Diamond74.6%
Rank 277 of 619 measured models
HLE11.4%
Rank 303 of 664 measured models
SciCode45.1%
Rank 143 of 231 measured models
AA-LCR58.7%
Rank 292 of 575 measured models
Terminal-Bench 2.156.2%
Rank 110 of 244 measured models
Terminal-Bench v40.5%
Rank 136 of 218 measured models
τ-Bench Banking15.7%
Rank 118 of 211 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