GPT-5.4 Nano
GPT-5.4 Nano is OpenAI's compact GPT-5.4-class model, built for classification, data extraction, and coding subagents, available only via the API.
Model Overview
Capabilities, design details, and architectural traits
GPT-5.4 Nano - OpenAI's GPT-5.4-Class Model
GPT-5.4 Nano is a model in OpenAI's GPT-5.4 family. It trades deep reasoning for speed, and OpenAI recommends it specifically for classification, data extraction, and coding subagents that handle simpler supporting tasks. Unlike its larger siblings, it is built to run at very high volume rather than to lead complex, multi-step work.
| Trait | Detail |
|---|---|
| Positioning | Built for simple, high-volume tasks |
| Recommended use | Classification, data extraction, and coding subagents handling simpler supporting tasks |
| Availability | API-only - not offered in ChatGPT or the Codex app, CLI, or IDE extension |
| Reasoning control | reasoning.effort can be set to none (default), low, medium, high, or xhigh |
| Computer-use tool | Not supported, unlike GPT-5.4 and GPT-5.4 mini, which both support computer-use tasks |
API-Only, No Consumer Surface
GPT-5.4 Nano has no presence in ChatGPT or Codex. OpenAI reserves it for direct API integration, while GPT-5.4 mini ships in ChatGPT's Free and Go tiers and across the Codex app, CLI, IDE extension, and web.
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | OpenAI | Anthropic | Anthropic |
| Release Date | March 17, 2026 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | Aug 2025 | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 400K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.20 Best Input Pricing | $4 | $2 |
| Output Pricing | $1.25 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | fileimagetext | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 20.0 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | - | - | - |
| Agentic Index | - | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from OpenAI
Other models by OpenAI
Top AI Models
Leading alternatives by intelligence score
