GPT-4 Turbo
by OpenAItext+image->text1 endpoint
Neura Intelligence Index
59.6/ 100
Rank
#135—
Confidence
highBenchmarks
12(31% coverage)
Overview
The latest GPT-4 Turbo model with vision capabilities. Vision requests can now use JSON mode and function calling.
Training data: up to December 2023.
Capabilities
Text generation
Image understanding
Tool use / Function calling
Structured output
Long context
Top-P sampling
Stop sequences
Deterministic seed
Log probabilities
Modalities
Input
TextImage
Output
Text
Technical Specifications
Context Window
128K tokens
Max Output
4.1K tokens
Knowledge Cutoff
2023-12-31
Tokenizer
GPT
Content Moderation
Enabled
Uptime (24h)
10000.0%
Blended Cost
$15/1M tokens
Country
🇺🇸US
Model Family
GPT-4
Release Date
April 9, 20242y
Supported Parameters
frequency_penaltylogit_biaslogprobsmax_tokenspresence_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_logprobstop_p
Benchmark Performance
general
| Benchmark | Score | Source |
|---|---|---|
| Chatbot Arena Elo | 1256 | aggregated |
| IFEval | 82.1 | aggregated |
| MMLU-Pro | 64.1 | aggregated |
coding
| Benchmark | Score | Source |
|---|---|---|
| HumanEval | 87.6 | aggregated |
reasoning
| Benchmark | Score | Source |
|---|---|---|
| ARC-Challenge | 96.3 | aggregated |
| BigBench-Hard | 85.1 | aggregated |
| GPQA Diamond | 49.3 | aggregated |
language
| Benchmark | Score | Source |
|---|---|---|
| HellaSwag | 95.3 | aggregated |
| WinoGrande | 87.5 | aggregated |
safety
| Benchmark | Score | Source |
|---|---|---|
| TruthfulQA | 70.1 | aggregated |