Free ChatGPT users now have access to near-flagship AI capabilities with OpenAI's new GPT-5.4 mini, one of two compact models launched this week that deliver responses more than twice as fast as their predecessors.
The GPT-5.4 mini and nano models target high-volume workloads where speed and cost matter more than maximum capability. Both inherit strengths from the full GPT-5.4 system but run with significantly lower latency and operating expenses.
GPT-5.4 mini brings improvements across coding, reasoning, multimodal understanding, and tool use while maintaining close performance to the larger flagship model. On the GPQA Diamond test measuring high-level expertise, it scored 88.01% compared to the flagship's 93%. For programming tasks on SWE-Bench Pro, it reached 54.4% versus 57.7% for GPT-5.4.
Available immediately through multiple channels, GPT-5.4 mini reaches free ChatGPT users via the "Thinking" feature in Free and Go tiers. Developers can access it through API, Codex app, command-line interface, and IDE extensions.
The smaller GPT-5.4 nano focuses on lightweight workloads like classification and data extraction at minimal cost. Priced at $0.20 per million input tokens compared to $2.50 for full GPT-5.4, it targets simple repetitive tasks where response time directly impacts user experience.
Both models represent a shift toward specialized AI deployment rather than one-size-fits-all solutions.
"In these settings, the best model is often not the largest one, it's the one that can respond quickly, use tools reliably, and still perform well on complex professional tasks," according to OpenAI's announcement.
The launch follows OpenAI's release of GPT-5.4 earlier this month, which introduced mid-response change and improved deep web research capabilities.
