GPT-5.4 nano
GPT-5.4 Nano is designed for tasks where speed and cost matter most like classification, data extraction, ranking, and sub-agents.
text + image + file → text · made by OpenAI
$0.2→$1.25per Mtok in / out
OpenAI · 11 sellers
Sold by 23 ways
| Seller | Lane | Rate |
|---|---|---|
| Microsoft Azure AI cloud | flex | $0.1per Mtok in$0.62per Mtok outraw.githubusercontent.com · read 2026-10-31 |
| Microsoft Azure AI cloud | batch | $0.11per Mtok in$0.69per Mtok out$0.01per Mtok cachedraw.githubusercontent.com · read 2026-10-31 |
| Databricks cloud | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cached$0.2per Mtok cache writemodels.dev · read 2026-10-31 |
| DigitalOcean cloud | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedmodels.dev · read 2026-10-31 |
| Microsoft Azure AI cloud | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedmodels.dev · read 2026-10-31 |
| OpenAI api | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cacheddevelopers.openai.com · read 2026-08-24 |
| Writer api | standard | $0.2per Mtok in$1.25per Mtok outdev.writer.com · read 2026-08-25 |
| Microsoft Azure AI cloud | priority | $0.4per Mtok in$2.5per Mtok out$0.04per Mtok cachedraw.githubusercontent.com · read 2026-10-31 |
| Requesty aggregatornot on the seller's list today | openai flex | $0.09per Mtok in$0.56per Mtok out$0.009per Mtok cachedrouter.requesty.ai · read 2026-08-25 |
| Requesty aggregatornot on the seller's list today | openai-responses flex | $0.09per Mtok in$0.56per Mtok out$0.009per Mtok cachedrouter.requesty.ai · read 2026-08-25 |
| Requesty aggregatornot on the seller's list today | openai-responses | $0.1per Mtok in$0.62per Mtok out$0.01per Mtok cachedrouter.requesty.ai · read 2026-09-10 |
| Requesty aggregator | openai | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedrouter.requesty.ai · read 2026-10-10 |
| Nous Research 2 ways | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cached | |
| Nous Research aggregator | batch | $0.1per Mtok in$0.62per Mtok out$0.01per Mtok cachedinference-api.nousresearch.com · read 2026-09-12 |
| Nous Research aggregator | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedinference-api.nousresearch.com · read 2026-09-12 |
| OpenRouter 4 ways | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cached | |
| OpenRouter aggregator | batch | $0.1per Mtok in$0.62per Mtok out$0.01per Mtok cachedopenrouter.ai · read 2026-08-24 |
| OpenRouter aggregator | openai/flex | $0.1per Mtok in$0.62per Mtok out$0.01per Mtok cachedopenrouter.ai · read 2026-08-26 |
| OpenRouter aggregator | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedopenrouter.ai · read 2026-08-24 |
| OpenRouter aggregator | azure/us | $0.22per Mtok in$1.38per Mtok out$0.022per Mtok cachedopenrouter.ai · read 2026-08-25 |
| AtlasCloud 0 ways | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cached | |
| AtlasCloud aggregatornot on the seller's list today | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedapi.atlascloud.ai · read 2026-09-20 |
| ElectronHub aggregator | standard | $0.2per Mtok in$1.25per Mtok outapi.electronhub.ai · read 2026-10-31 |
| GitHub Copilot aggregator | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedmodels.dev · read 2026-10-31 |
| Vercel AI Gateway aggregator | standard | $0.2per Mtok in$1.25per Mtok out$0.02per Mtok cachedvercel.com · read 2026-08-25 |
| Vercel AI Gateway aggregatornot on the seller's list today | sandbox | $0.2per Mtok in$1.25per Mtok outvercel.com · read 2026-08-25 |
Measured 28 standings
| Place | Board | Metric | Score |
|---|---|---|---|
| 2ndof 111 | Vectara Hallucination Leaderboard | Hallucination Rate | 3.1 |
| 29thof 154 | Chess puzzles — Epoch AI | mean_score (xhigh) | 45 |
| 29thof 72 | Frontiermath tier 4 — Epoch AI | mean_score (high) | 6.25 |
| 33rdof 101 | FrontierMath (Tiers 1-3) | mean_score (high) | 25.86 |
| 61stof 122 | Frontiermath tiers 1 3 v2 — Epoch AI | mean_score (high) | 44.912 |
| 62ndof 132 | Ale bench — Epoch AI | Performance (high) | 1004.52 |
| 65thof 163 | Artificial Analysis · Coding Index | Index | 56.1 |
| 66thof 78 | Frontiermath tier 4 v2 — Epoch AI | mean_score (high) | 12.195 |
| 73rdof 131 | Artificial Analysis · Agentic Index | Index | 16 |
| 74thof 173 | Weirdml — Epoch AI | Accuracy (high) | 49.23 |
| 76thof 84 | Proofbench — Epoch AI | Accuracy (high) | 5 |
| 86thof 92 | Forecastbench — Epoch AI | Overall score | 58.6 |
| 88thof 306 | OTIS Mock AIME 2024-2025 | mean_score (high) | 87.78 |
| 89thof 93 | Simpleqa verified — Epoch AI | mean_score (high) | 11.7 |
| 93rdof 165 | LMArena · Vision | Score (Elo) (high) | 1199 |
| 119thof 142 | Mystery game puzzles — Epoch AI | mean_score (none) | 9 |
| 122ndof 260 | Critpt — Epoch AI | Accuracy (xhigh) | 9.253 |
| 123rdof 197 | Lmca — Epoch AI | Score (xhigh) | 36.86 |
| 133rdof 235 | Dtbench — Epoch AI | Accuracy (xhigh) | 80.27 |
| 139thof 235 | Scicode — Epoch AI | Score (xhigh) | 47.222 |
| 144thof 326 | GPQA Diamond | mean_score (high) | 78.47 |
| 159thof 414 | LMArena · Text | Score (Elo) (high) | 1401 |
| 178thof 282 | ARC-AGI-2 (Public Eval) | Score (%) (high) | 5.14 |
| 190thof 298 | ARC-AGI-2 (Semi-Private) | Score (%) (xhigh) | 5.69 |
| 191stof 290 | Arc agi 2 — Epoch AI | Score (xhigh) | 5.69 |
| 192ndof 273 | ARC-AGI-1 (Public Eval) | Score (%) (high) | 51.62 |
| 192ndof 298 | ARC-AGI-1 (Semi-Private) | Score (%) (xhigh) | 51.5 |
| 202ndof 309 | Arc agi — Epoch AI | Score (xhigh) | 51.5 |
About
GPT-5.4 nano — a text model from OpenAI, sold by 11 companies from $0.2 in and $1.25 out per million tokens, placed 2nd of 111 on Vectara Hallucination Leaderboard.
It takes text, images and files and returns text, with a context window of 400,000 tokens. It was published in March 2026, trained on material up to August 2025. Its sellers say it can reason step by step and call a tool. The catalogue files it under extract, reasoning and agents. Eleven companies sell it. The cheapest is $0.2 in and $1.25 out per million tokens at Databricks. Beside the standard rate there are batch, flex, faster, regional and separately routed lanes. It has been measured on 28 boards, and stands best at 2nd of 111 on Vectara Hallucination Leaderboard.
Every current figure
- Maker
- OpenAI
- Register
- model
- Takes
- text + image + file
- Returns
- text
- Context
- 400,000 tokens
- Longest answer
- 128,000 tokens
- Published
- March 2026
- Knowledge to
- August 2025
- Licence
- not read
- Sellers
- 11
- Price
- $0.2 in and $1.25 out per million tokens — Databricks
- Boards
- 28
- Best place
- 2nd of 111 — Vectara Hallucination Leaderboard
Known as 9 names
openai/gpt-5.4-nano5.4 nanoopenai/gpt-5.4-nano:batchgpt-5.4-nanoopenai-responses/gpt-5.4-nanoopenai-responses/gpt-5.4-nano:flexopenai/gpt-5.4-nano:flexopenai/gpt-5.4-nano-batchOpenAI: GPT-5.4 Nano (batch)