OpenAI released GPT-5.5 on April 23, 2026, marking a significant advancement in artificial intelligence capabilities. The GPT-5.5 release introduces a model designed to handle complex, multi-step tasks with greater autonomy and reasoning ability than its predecessor.

The model excels at writing and debugging code, researching online, analyzing data, creating documents and spreadsheets, operating software, and moving across tools to complete tasks. GPT-5.5 understands user intent faster and carries more of the work itself, reducing the need for constant manual intervention.

Performance and Speed Improvements

GPT-5.5 delivers performance gains without sacrificing speed. The model matches GPT-5.4 per-token latency in real-world serving while performing at a significantly higher intelligence level. Notably, GPT-5.5 uses fewer tokens to complete the same tasks, making it more efficient as well as more capable.

On Terminal-Bench 2.0, which tests complex command-line workflows, GPT-5.5 achieves 82.7% accuracy. On SWE-Bench Pro, evaluating real-world GitHub issue resolution, it reaches 58.6%, solving more tasks end-to-end in a single pass than previous models. The model also outperforms GPT-5.4 on Expert-SWE, an internal evaluation for long-horizon coding tasks.

Agentic Coding Capabilities

Early testers reported that GPT-5.5 shows stronger ability to understand system architecture, identify failure points, and determine where fixes should land. Dan Shipper, founder and CEO of Every, described GPT-5.5 as “the first coding model I’ve used that has serious conceptual clarity.” He tested the model by asking it to rewrite part of a system after a post-launch issue; GPT-5.4 could not produce the same rewrite that a senior engineer eventually decided on, but GPT-5.5 could.

Pietro Schirano, CEO of MagicPath, observed a similar step change when GPT-5.5 merged a branch with hundreds of frontend and refactor changes into a substantially modified main branch, resolving the work in one shot in approximately 20 minutes. Senior engineers noted that GPT-5.5 was noticeably stronger than GPT-5.4 and Claude Opus 4.7 at reasoning and autonomy, catching issues in advance and predicting testing and review needs without explicit prompting.

Knowledge Work and Professional Applications

The same strengths that make GPT-5.5 effective for coding also apply to everyday work on computers. The model better understands intent and moves more naturally through the full loop of knowledge work: finding information, understanding what matters, using tools, checking output, and turning raw material into something useful.

In Codex, GPT-5.5 outperforms GPT-5.4 at generating documents, spreadsheets, and slide presentations. Alpha testers reported superior performance on operational research, spreadsheet modeling, and converting messy business inputs into plans. More than 85% of OpenAI employees use Codex weekly across functions including software engineering, finance, communications, marketing, data science, and product management.

OpenAI’s Finance team used Codex to review 24,771 K-1 tax forms totaling 71,637 pages, accelerating the task by two weeks compared to the prior year. The Go-to-Market team automated generating weekly business reports, saving 5-10 hours per week. On GDPval, which tests agents’ abilities to produce well-specified knowledge work across 44 occupations, GPT-5.5 scores 84.9%.

Scientific Research Applications

GPT-5.5 shows gains on scientific and technical research workflows that require exploration, evidence gathering, assumption testing, result interpretation, and decision-making about next steps. The model demonstrates clear improvement over GPT-5.4 on GeneBench, a new evaluation focusing on multi-stage scientific data analysis in genetics and quantitative biology.

An internal version of GPT-5.5 helped discover a new proof about Ramsey numbers, a central object in combinatorics. The model found a proof of a longstanding asymptotic fact about off-diagonal Ramsey numbers, later verified in Lean. Derya Unutmaz, an immunology professor at the Jackson Laboratory for Genomic Medicine, used GPT-5.5 Pro to analyze a gene-expression dataset with 62 samples and nearly 28,000 genes, producing a detailed research report that would have taken his team months.

Safety and Cybersecurity Measures

OpenAI released GPT-5.5 with what the company describes as its strongest set of safeguards to date. The model was evaluated across the company’s full suite of safety and preparedness frameworks, with input from internal and external red teamers. Targeted testing covered advanced cybersecurity and biology capabilities, and feedback was collected from nearly 200 trusted early-access partners before release.

GPT-5.5 is treated as High under OpenAI’s Preparedness Framework for biological, chemical, and cybersecurity capabilities. On Capture-the-Flags challenge tasks, GPT-5.5 achieves 88.1% accuracy, and on CyberGym, it reaches 81.8%. OpenAI deployed stricter classifiers for potential cyber risk and designed tighter controls around higher-risk activity and sensitive cyber requests.

The company is making cyber-permissive models available through Trusted Access for Cyber, starting with Codex. Organizations responsible for defending critical infrastructure can apply to access cyber-permissive models while meeting strict security requirements. Users can apply for trusted access at chatgpt.com/cyber to reduce unnecessary refusals while using GPT-5.5 for verified defensive work.

Availability and Pricing

GPT-5.5 is rolling out to Plus, Pro, Business, and Enterprise users in ChatGPT and Codex. GPT-5.5 Pro is rolling out to Pro, Business, and Enterprise users in ChatGPT. In Codex, GPT-5.5 is available for Plus, Pro, Business, Enterprise, Edu, and Go plans with a 400K context window. GPT-5.5 is also available in Fast mode, generating tokens 1.5x faster for 2.5x the cost.

For API developers, gpt-5.5 will be available in the Responses and Chat Completions APIs at $5 per 1M input tokens and $30 per 1M output tokens, with a 1M context window. Batch and Flex pricing are available at half the standard API rate, while Priority processing is available at 2.5x the standard rate. GPT-5.5-Pro will be priced at $30 per 1M input tokens and $180 per 1M output tokens. While GPT-5.5 is priced higher than GPT-5.4, it is both more intelligent and significantly more token efficient.

Source: OpenAI