OpenAI logo

GPT-5.5 Instant

Multimodal
OpenAI

GPT-5.5 Instant is OpenAI's latest Instant model and the default ChatGPT model, released on May 5, 2026. It improves factuality, everyday reasoning, image understanding, STEM answers, concision, conversational tone, and personalization over GPT-5.3 Instant, and is exposed through the chat-latest alias in the API.

Key Specifications

Parameters
-
Context
400.0K
Release Date
May 5, 2026
Average Score
68.8%

Timeline

Key dates in the model's history
Announcement
May 5, 2026
Today
August 27, 2026

Technical Specifications

Parameters
-
Training Tokens
-
Knowledge Cutoff
August 31, 2025
Family
-
Capabilities
MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)
$5.00
Output (per 1M tokens)
$30.00
Max Input Tokens
400.0K
Max Output Tokens
128.0K
Supported Features
Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Reasoning

Logical reasoning and analysis
GPQA
GPQA PhD-level science accuracy.Self-reported
85.6%

Other Tests

Specialized benchmarks
AIME 2025
AIME 2025 competition math accuracy.Self-reported
81.2%
CharXiv-R
CharXiv reasoning scientific chart reasoning accuracy.Self-reported
81.6%
HealthBench
Length-adjusted HealthBench score. Unadjusted score 50.9; mean response length 1,922 characters.Self-reported
51.4%
HealthBench Consensus
Length-adjusted HealthBench Consensus score. Unadjusted score 94.6; mean response length 1,919 characters.Self-reported
94.7%
HealthBench Hard
Length-adjusted HealthBench Hard score. Unadjusted score 21.3; mean response length 1,794 characters.Self-reported
22.9%
HealthBench Professional
Length-adjusted HealthBench Professional score. Unadjusted score 40.7; mean response length 2,775 characters.Self-reported
38.4%
MMMU-Pro
MMMU-Pro expert multimodal reasoning accuracy.Self-reported
76.0%
OmniDocBench 1.5
OmniDocBench document parsing average error rate 0.125 inverted to normalized score 0.875 (1 - 0.125).Self-reported
87.5%

License & Metadata

License
proprietary
Announcement Date
May 5, 2026
Last Updated
August 27, 2026

Similar Models

All Models

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.