OpenAI Proprietary Released May 5, 2026 Updated
GPT-5.5 Instant
ToolsGuide Scores, rankings, test results and price.
Model summary
Summary
GPT-5.5 Instant is OpenAI's model, released May 5, 2026. It's ranked in 2 categories and does best at vision: #27 of 167, with a ToolsGuide Score of 540 against a median of 346.
It also ranks #84 of 416 for text (405).
It costs $5.00 per million input tokens and $30.00 per million output tokens, about 4.0× the median vision model ($1.25).
It has published results on 9 tests.
Specifications
- Creator
- OpenAI
- Licence
- Proprietary
- Released
- May 5, 2026
- Context window
- 400K tokens
- Input price
- $5.00 / 1M
- Output price
- $30.00 / 1M
- Tests reported
- 9
- Settings ranked
- gpt-5.5-instant, GPT-5.5 Instant
Model comparison
Against the ten best vision models.
Vision score
ToolsGuide Score out of 1,000 · Higher is better
Text score
ToolsGuide Score out of 1,000 · Higher is better
Input price
USD per 1M input tokens · Lower is better
Leaderboards
Where GPT-5.5 Instant sits among the top 20 in each category.
Text
#84 of 416 · ToolsGuide Score out of 1,000, where 1,000 is a flawless model
Vision
#27 of 167 · ToolsGuide Score out of 1,000, where 1,000 is a flawless model
By task
Scored from votes on each kind of prompt alone, on the same 1,000-point scale.
Score against price
Vision · the 10 priced models ranked closest · up and to the left is better value
Against its nearest rivals
Text prompts by kind · ToolsGuide Score from votes · hover for values
- GPT-5.5 Instant
-
Claude Sonnet 4.6 -
Gemini 3.6 Flash (High)
Text
gpt-5.5-instant · hover a bar for its rank
Vision
GPT-5.5 Instant · hover a bar for its rank
Test results
Every test GPT-5.5 Instant has a published result on, out of 100%, with its rank among the models that took it.
GPQA
#66 / 256Graduate-level science questions written to be hard to look up.
85.6%
Self-reported
AIME 2025
#73 / 123Competition maths problems from the 2025 American Invitational Mathematics Exam.
81.2%
Self-reported
MMMU-Pro
#30 / 73College-level questions that need both an image and text to answer.
76.0%
Self-reported · In the vision score
CharXiv-R
#22 / 60Reading and reasoning about charts from scientific papers.
81.6%
Self-reported
HealthBench Hard
#13 / 24A challenging variation of HealthBench that evaluates large language models' performance and safety in healthcare through 5,000 multi-turn conversations with particularly rigorous evaluation criteria validated by 262 physicians from 60 countries
22.9%
Self-reported
OmniDocBench 1.5
#11 / 20OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.
87.5%
Self-reported
HealthBench Professional
#18 / 19HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.
38.4%
Self-reported
HealthBench
#12 / 14An open-source benchmark for measuring performance and safety of large language models in healthcare, consisting of 5,000 multi-turn conversations evaluated by 262 physicians using 48,562 unique rubric criteria across health contexts and behavioral dimensions
51.4%
Self-reported
HealthBench Consensus
#5 / 5HealthBench Consensus is a HealthBench subset focused on questions where physician-created rubric criteria have especially high agreement, measuring healthcare performance and safety on consensus-evaluable conversations.
94.7%
Self-reported