OpenAI model product
GPT-5.4-nano
4-class model for simple high-volume tasks.
Updated Aug 12, 2026. Default version: GPT-5.4 nano
LLMBoard score49.1GPT-5.4 nano
Coverage60%15 benchmark families
Context window400KTokens
Official input price$0.20OpenAI API
GPT-5.4-nano Specifications
Technical details for the model's default version.
- Version
- GPT-5.4 nano
- Released
- Mar 17, 2026
- Knowledge cutoff
- Aug 31, 2025
- Parameters
- N/A
- Context window
- 400K
- Max output
- 128K
- Inputs
- image, text
- Outputs
- text
- Open weights
- No
- License
- Proprietary
GPT-5.4-nano Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
GPT-5.4-nano Benchmark Results
Benchmark scores for GPT-5.4 nano.
| Graphwalks BFS <128k | 73.4% | 05 | 11 | 60.0% | C | |
| Graphwalks parents <128k | 50.8% | 09 | 11 | 20.0% | C | |
| MRCR v2 (8-needle) | 33.1% | 12 | 21 | 45.0% | C | |
| OmniDocBench 1.5 | 75.8% | 12 | 17 | 31.3% | C | |
| Tau2 Telecom | 92.5% | 15 | 35 | 58.8% | C | |
| Finance Agent v2 | 38.2% | 17 | 26 | 36.0% | B | |
| OSWorld-Verified | 39.0% | 23 | 23 | 0.0% | C | |
| Toolathlon | 35.5% | 27 | 31 | 13.3% | C | |
| LiveBench | 70.1% | 30 | 38 | 21.6% | B | |
| MCP Atlas | 56.1% | 30 | 31 | 3.3% | C | |
| SWE-Bench Pro | 52.4% | 39 | 45 | 13.6% | C | |
| MMMU-Pro | 66.1% | 41 | 66 | 38.5% | C | |
| Terminal-Bench 2.0 | 46.3% | 41 | 49 | 16.7% | C | |
| Humanity's Last Exam | 24.3% | 51 | 93 | 45.6% | C | |
| GPQA | 82.8% | 70 | 234 | 70.4% | C |
GPT-5.4-nano Arena Results
Preference and agent-evaluation results for the default version.
| vision style control | overall | 70 | 1202.4 | 23,332 | N/A | |
| vision | overall | 73 | 1198.5 | 23,332 | N/A | |
| text factuality | overall | 108 | 1420.0 | 58,530 | N/A | |
| text style control | overall | 134 | 1402.3 | 58,654 | N/A | |
| text | overall | 155 | 1372.9 | 58,654 | N/A |
GPT-5.4-nano Runtime Performance
Provider-specific output speed and catalog latency for GPT-5.4 nano. Runtime does not affect the capability score.
| OpenAI | 234.241 tok/s | 2.384 s | 400K | 128K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
GPT-5.4-nano Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.20 input, $1.25 output per 1M
- Official provider
- OpenAI
- Lowest third-party
- From $0.18 input, $1.1 output per 1M via Poe
- Tracked offerings
- 27
| Poe | openai/gpt-5.4-nano | global | $0.18 | $1.1 | 400K | |
| Impossibl | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| CrossModel | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Vivgrid | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| NanoGPT | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| FastRouter | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| NEAR AI Cloud | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| LLM Gateway | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| OpenCode Zen | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| GitHub Copilot | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| 302.AI | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Neon | gpt-5-4-nano | global | $0.20 | $1.25 | 400K | |
| FrogBot | gpt-5-4-nano | global | $0.20 | $1.25 | 400K | |
| Pioneer | gpt-5.4-nano | global | $0.20 | $1.25 | 1M | |
| ZenMux | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Abacus | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Azure Cognitive Services | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Azure | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Ofox | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| OrcaRouter | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Kilo Gateway | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Merge Gateway | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Vercel AI Gateway | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| OpenRouter | openai/gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| OpenAI | gpt-5.4-nano | global | $0.20 | $1.25 | 400K | |
| Databricks | databricks-gpt-5-4-nano | global | $0.20 | $1.25 | 400K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
GPT-5.4-nano Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| GPT-5.4 nano | 49.1 | N/A | 400K | 128K | No | Proprietary |
What is GPT-5.4-nano?
Key information about GPT-5.4-nano and its available data.
4-class model for simple high-volume tasks. It is optimized for classification, data extraction, ranking, and lightweight coding subagents, while preserving strong reasoning, multimodal understanding, and tool use for latency-sensitive workloads.
Data as of 2026-08-11.
GPT-5.4-nano vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to GPT-5.4-nano
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
OP
GPT-5.1-Codex
OpenAI
OP
GPT-5
OpenAI
OP
GPT-OSS-20B
OpenAI
OP
o3
OpenAI
OP
o4 mini
OpenAI
OP
GPT-5-mini
OpenAI
FAQ
Common questions about GPT-5.4-nano.
When was GPT-5.4-nano released?
GPT-5.4-nano's default version was released on Mar 17, 2026.
How much does GPT-5.4-nano cost?
GPT-5.4-nano's official API price is $0.20 per million input tokens and $1.25 per million output tokens via OpenAI. The lowest tracked third-party offer starts at $0.18 input and $1.1 output via Poe.
Who created GPT-5.4-nano?
GPT-5.4-nano was created by OpenAI.
What is the context window for GPT-5.4-nano?
The default version has a 400K token context window.
Is GPT-5.4-nano open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer GPT-5.4-nano?
27 provider offerings are linked to the default version.
What models should I compare GPT-5.4-nano with?
Nearby ranked alternatives include GPT-5.1-Codex, GLM 4.5, Mistral Medium 3.5.