मुख्य सामग्री पर जाएं

GPT-5.6 Terra (max)

OpenAIGPTProprietary

विवरण

GPT-5.6 Terra is the balanced tier of OpenAI's GPT-5.6 family, designed for workloads that balance intelligence and cost. It delivers performance competitive with GPT-5.5 at roughly half the price, supports max reasoning effort, and has a 1.05M-token context window.

रिलीज़ तिथि
2026-07-09
पैरामीटर
संदर्भ लंबाई
1.1M
मोडैलिटीज़
image, pdf, text

क्षमता रडार

54
general
73
coding
93
reasoning
69
science
70
agents
85
multimodal

रैंकिंग

बेंचमार्क स्कोर (LLM Stats)

(LLM Stats (zeroeval))

Agents

Connectors100.0%स्वयं
Search and Function-Calling94.6%स्वयं
Capture-the-Flag Challenges (Internal)91.8%स्वयं
Internal Research Debugging Evaluation67.8%स्वयं
SEC-bench Pro57.7%स्वयं
RSI Index56.3%स्वयं
LifeSciBench56.0%स्वयं
Toolathlon53.1%स्वयं
PostTrainBench Lite51.5%स्वयं
Big Finance Bench51.0%स्वयं
Agents' Last Exam50.4%स्वयं
KernelGen 1P49.2%स्वयं
Management Consulting Tasks (Internal)37.2%स्वयं
GeneBench-Pro23.3%स्वयं
ExploitGym23.2%स्वयं
AutomationBench15.2%स्वयं
NanoGPT14.5%स्वयं

Code

BenchCAD (with Python tool)78.2%स्वयं
DeepSWE 1.170.0%
DeepSWE69.6%स्वयं
BenchCAD62.3%स्वयं
ExploitBench52.9%स्वयं

General

Artificial Analysis55.0%
GDP.pdf24.7%स्वयं

Healthcare

HealthBench Consensus95.1%स्वयं
HealthBench Professional57.7%स्वयं
HealthBench57.0%स्वयं
HealthBench Hard32.7%स्वयं

Long Context

MRCR v2 (8-needle)89.6%स्वयं
MRCR v2 (8-needle, 512K-1M)72.5%स्वयं

Math

FrontierMath84.9%स्वयं
FrontierMath Tier 4 (v2)68.3%स्वयं

Multimodal

OSWorld 2.050.2%स्वयं

Reasoning

GPQANYU + Cohere + Anthropic (2023)92.9%स्वयं
BrowseCompOpenAI (2025)87.5%स्वयं
Terminal-Bench 2.187.4%स्वयं
Graphwalks BFS >128k76.9%स्वयं
Graphwalks BFS 1M71.2%स्वयं
SWE-Bench ProPrinceton NLP (2024)63.4%स्वयं
FrontierCode 1.141.3%
MedChemBench (Internal)35.0%स्वयं
ARC-AGI-30.8%स्वयं

Vision

MMMU-Pro (with tools)82.0%स्वयं
MMMU-Pro80.7%स्वयं

AA मूल्यांकन सूचकांक

(Artificial Analysis)
Coding Index(Artificial Analysis)
76.7
Intelligence Index(Artificial Analysis)
56.6
Gpqa(NYU + Cohere + Anthropic (2023))
0.9
Terminalbench V2 1
0.9
Tau2(Sierra + U Toronto + Vector Institute (2025))
0.9
Lcr(Artificial Analysis)
0.8
Ifbench(Google Research (2023))
0.7
Terminalbench Hard(Stanford × Laude Institute (2026))
0.6
Scicode(UIUC + Argonne National Lab (2024))
0.5
Hle(Center for AI Safety + Scale AI (2025))
0.4
Tau Banking
0.4

LLM Stats श्रेणी स्कोर

(LLM Stats (zeroeval))
Physics
90
Search
90
Biology
90
Chemistry
90
Long Context
80
Math
80
Spatial Reasoning
70
Tool Calling
70
Multimodal
60
Reasoning
60
Safety
60
General
60
Healthcare
60
Agents
60
Code
60
Vision
60
Science
40
Finance
40
Systems
40

मूल्य निर्धारण

इनपुट मूल्य$2 / 1M टोकन
आउटपुट मूल्य$12 / 1M टोकन
मिश्रित मूल्य (3:1)$4.5 / 1M टोकन
कैश पठन मूल्य$0.2 / 1M टोकन
कैश लेखन मूल्य$2.5 / 1M टोकन

गति

टोकन/सेकंड113.9
पहले टोकन में देरी142.73s
पहले उत्तर में देरी142.73s

प्रदाता मूल्य रैंकिंग

प्रदाता मूल्य रैंकिंग

2 प्रदाता

सबसे सस्ता: OpenAIसबसे महंगा: Neon
प्रदाताइनपुटआउटपुट
1OpenAIसबसे सस्ता
$0
$0.00001
2Neon
$2.5
$15

इस मॉडल के लिए विभिन्न API प्रदाताओं के मूल्य निर्धारण की तुलना करें।

बाहरी लिंक