GPT-5.3 Instant System Card

           OpenAI

        March 3, 2026

             1
1       Introduction

GPT-5.3 Instant is the newest addition to the GPT-5 series. As described in our blog, GPT-5.3
Instant responds faster, delivers richer and better-contextualized answers when searching the web,
and reduces unnecessary dead ends, caveats, and overly declarative phrasing that can interrupt
the flow of conversation. The comprehensive safety mitigation approach for this model is largely
the same as that described for GPT-5.2 Instant in the GPT-5.2 System Card.

In this card we also refer to GPT-5.3 Instant as gpt-5.3-instant.

2       Model Data and Training

Like OpenAI’s other models, this model was trained on diverse datasets, including information
that is publicly available on the internet, information that we partner with third parties to access,
and information that our users or human trainers and researchers provide or generate. Our data
processing pipeline includes rigorous filtering to maintain data quality and mitigate potential
risks. We use advanced data filtering processes to reduce personal information from training data.
We also employ safety classifiers to help prevent or reduce the use of harmful or sensitive content,
including explicit materials such as sexual content involving a minor.

Note that we are continually iterating on our models. Evaluations in this card for GPT-5.3 Instant
are for the version shipped on 2/26/2026. Similarly, comparison values from previously-launched
models are from the latest versions of those models as of launching GPT-5.3 Instant, so may vary
slightly from values published in previous cards.1

3       Safety

3.1     Disallowed Content

We conducted benchmark evaluations across disallowed content categories. We report here on
our Production Benchmarks, an evaluation set with conversations representative of challenging
examples from production data. As we noted in previous system cards, we introduced these
Production Benchmarks to help us measure continuing progress given that our earlier Standard
evaluations for these categories had become relatively saturated.

These evaluations were deliberately created to be difficult. They were built around cases in
which our existing models were not yet giving ideal responses, and this is reflected in the scores
below. Error rates are not representative of average production traffic. The metric is not_unsafe,
checking that the model did not produce output that is disallowed under the relevant OpenAI
policy.

Values from previously launched models are from the latest versions of those models, and are
    1
     GPT-5.3 Instant is intended to be used in accordance with OpenAI’s Usage Policies, Service Terms, and Terms
of Use. These policies apply universally to OpenAI services and are designed to ensure safe and responsible usage
of AI technology. You can review OpenAI’s Usage Policies at openai.com/policies/usage-policies/ .
   If you need assistance with respect to GPT-5.3 Instant, you can find further information on OpenAI’s website
(openai.com), or you can contact OpenAI Support by opening the chat bubble icon displayed at the bottom-right
of help.openai.com.

                                                       1
subject to some variation. Values may vary slightly from values published at launch for those
models.

                         Table 1: Production Benchmarks (higher is better)

            Content type            gpt-5.1-instant      gpt-5.2-instant      gpt-5.3-instant
            violent illicit              0.962                0.965                0.926
            behavior
            nonviolent illicit           0.656                0.832                0.921
            behavior
            Self-harm                    0.874                0.923                0.895
            biology                      1.000                1.00                 1.00
            sexual content               0.930                0.926                0.866
            extremism                    1.000                0.979                0.981
            (propaganda praise
            or assistance)
            hate (abusive                0.959                0.851                0.868
            conduct protected
            class)
            graphic violence or          0.889                0.852                0.781
            physical injury

On average, the model performs above gpt-5.1-instant and below gpt-5.2-instant on our disallowed
content evaluations. gpt-5.3-instant shows regressions relative to gpt-5.2-instant and gpt-5.1-
instant for disallowed sexual content, and relative to gpt-5.2-instant for self-harm on both standard
and dynamic evaluations. The regressions for graphic violence and violent illicit behavior have
low statistical significance. For other categories, gpt-5.3-instant is either on par or improves upon
previous launches.

In addition to the benchmark evaluations, we reviewed safety during online experimentation and
did not observe an increase in undesirable responses for self-harm. We will continue monitoring
after launch to verify our test results and investigate the disparities between our offline evaluations
and online testing.

For disallowed sexual content, we deploy system-level safeguards in ChatGPT intended to mitigate
this behavior. We are continuing to improve our safeguards in this area and these learnings will
inform any future releases.

                              Table 2: Dynamic Mental Health Evaluations

      Evaluation                          gpt-5.1-instant     gpt-5.2-instant      gpt-5.3-instant
      mental health*                             0.832                1.000                0.985
      emotional reliance*                        0.945                0.952                0.992
      Self-harm *                                0.845                0.920                0.911

*Ahead of the GPT-5.3 Instant launch, we implemented dynamic multi-turn evaluations for
mental health, emotional reliance, and self-harm that simulate extended conversations across
these domains. Rather than assessing a single response within a fixed dialogue, these evaluations

                                                    2
allow conversations to evolve in response to the model’s outputs, creating varied trajectories
during testing that better reflect real user interactions. This approach helps identify potential
issues that may only emerge over the course of long exchanges and provides an even more rigorous
test than prior static multi-turn methods. By utilizing realistic, yet adversarial user simulations,
these evaluations have enabled continued improvements in safety performance, particularly in
areas where earlier evaluation frameworks had reached saturation.

Our standard evaluations measure whether the final model response violates our policies. In these
dynamic conversations, we instead evaluate whether any assistant response violates policy and
report the percentage of policy-compliant responses. The metric used is not_unsafe, representing
the share of assistant messages that do not violate safety policies.

4     Health Performance

4.1   HealthBench

Chatbots can empower consumers to better understand their health and help health professionals
deliver better care [1] [2]. We evaluate GPT-5.3 on HealthBench [3], an evaluation of health
performance and safety. HealthBench comprises 5,000 realistic (potentially multi-turn) health
conversations. Model responses are evaluated with example-specific rubrics. We report results on
three variants, HealthBench, HealthBench Hard, and HealthBench Consensus.

                                      Table 3: HealthBench

           Metric                  gpt-5.2-instant             gpt-5.3-instant
           HealthBench             55.4%                       54.1%
           Hard                    26.8%                       25.9%
           Consensus               95.8%                       95.3%
           Length                  2101 chars                  2140 chars

The major wins and losses below refer only to consensus-criteria deltas greater than 2.0% versus
GPT-5.2 instant.

Relative to GPT-5.2 instant, GPT-5.3 instant scores 54.1% on HealthBench (-1.3%), 25.9% on
Hard (-0.9%), and 95.3% on Consensus (-0.5%); at 2140 chars on average (+1.9%), it has slightly
worse performance at essentially matched length. On consensus criteria, its main strengths
are better context-seeking when important information is missing (+4.4%) and better hedging
behavior in irreducible-uncertainty settings (+4.0%). On consensus criteria, its main weaknesses
are poorer behavior in seeking context before referral (-10.1%), and lower accuracy when local
healthcare context may be pertinent but is not obvious (-5.5%).

References

[1] OpenAI, “Introducing gpt-5,” Aug. 2025. Accessed: 2025-12-10.

                                                 3
[2] OpenAI, “Pioneering an AI clinical copilot with Penda health,” July 2025. Accessed: 2025-12-
    10.

[3] OpenAI, “Introducing healthbench,” May 2025. Accessed: 2025-12-10.

                                               4