OpenAI Introduces GPT-5.4 Mini and Nano
OpenAI has officially launched GPT-5.4 mini and GPT-5.4 nano, introducing high-efficiency models optimized for automated workflows, coding subagents, and latency-sensitive deployments.
These models are designed to reduce application programming interface (API) overhead while maintaining complex reasoning capabilities, making them highly relevant for professionals scaling automated data extraction or telemetry analysis pipelines.
GPT-5.4 mini represents a significant architectural upgrade over its predecessor, delivering more than twice the execution speed of GPT-5 mini while approaching the performance metrics of the full GPT-5.4 model.
The model seamlessly handles text and image inputs, function calling, web search, file search, and direct computer use execution.
It operates with a massive 400,000-token context window, allowing sufficient capacity for parsing extensive system logs, analyzing dense user interface screenshots, and executing real-time multimodal reasoning.
GPT-5.4 nano is positioned as the most lightweight and cost-effective model in the OpenAI lineup. It is explicitly engineered for massive scale, prioritizing operational speed for workflows that do not require deep logical reasoning.
OpenAI recommends the nano variant for foundational tasks such as rapid classification, structured data extraction, ranking, and executing simpler supporting tasks within larger automated systems.
In rigorous technical evaluations, the mini model demonstrates strong capability retention despite its reduced computational footprint, maintaining high execution pass rates across software engineering and system terminal benchmarks.
| Benchmark Metric | GPT-5.4 (xhigh) | GPT-5.4 Mini (xhigh) | GPT-5.4 Nano (xhigh) | GPT-5 Mini (high) |
|---|---|---|---|---|
| SWE-Bench Pro (Public) | 57.7% | 54.4% | 52.4% | 45.7% |
| Terminal-Bench 2.0 | 75.1% | 60.0% | 46.3% | 38.2% |
| OSWorld-Verified | 75.0% | 72.1% | 39.0% | 42.0% |
| GPQA Diamond | 93.0% | 88.0% | 82.8% | 81.6% |
| Toolathlon | 54.6% | 42.9% | 35.5% | 26.9% |
The introduction of these models heavily emphasizes multi-agent system architectures, allowing developers to orchestrate complex execution pipelines by combining models of varying sizes.
Within environments like Codex, a primary larger model can manage overarching execution planning, system coordination, and final logical judgment.
Simultaneously, it can delegate narrower, parallelized tasks; these include searching codebases, reviewing massive configuration files, or processing supporting intelligence documents using GPT-5.4 mini subagents.
This tiered operational approach optimizes both computational latency and resource costs for extensive code debugging and vulnerability analysis loops.
GPT-5.4 mini is available immediately across the API, Codex, and ChatGPT, while GPT-5.4 nano is restricted exclusively to direct API access.
Within Codex environments, the mini model consumes only 30% of the standard GPT-5.4 allocation quota, substantially lowering the financial barrier for continuous automated analysis.
For standard consumer access, Free and Go tier users can utilize GPT-5.4 mini via the internal Thinking feature, while premium users will access it automatically as a rate limit fallback.
| Model Tier | Input Cost (per 1M tokens) | Output Cost (per 1M tokens) | Access Availability |
|---|---|---|---|
| GPT-5.4 Mini | $0.75 | $4.50 | API, Codex, ChatGPT |
| GPT-5.4 Nano | $0.20 | $1.25 | API Only |
Follow us on Google News, LinkedIn, and X to Get Instant Updates and Set GBH as a Preferred Source in Google.
Attackers abused compromised ConnectWise Cloud Backup credentials to access Microsoft 365 customer environments, prompting ConnectWise…
Anthropic says an internal Claude research model reconstructed how the National Security Agency generated seeds…
Two WordPress Core vulnerabilities can be chained to execute remote code without authentication, according to…
A newly disclosed OpenBSD kernel exploitation chain combines a shared memory accounting flaw with a…
GhostAction attackers have expanded their GitHub credential theft campaign with a second stage that disguises…
Indian authorities have warned that SideCopy hackers are targeting Indian organizations, including academic institutions, with…