Скорость вывода
19170.0 tok/s
20169.0 tok/s
22GPT-5.4 Mini165.0 tok/s
23162.3 tok/s
GPT-5.4 Mini by OpenAI is available through 217 API providers on LMSpeed. Compare API pricing from $0.0002 to $7492.50 per million input tokens across providers. Free API access is offered by 3 providers. Context window: 400,000. In speed benchmarks, the fastest provider reaches 245 tok/s.
OpenAI GPT-5.4 Mini is a compact language model in the GPT-5 series, optimized for quick responses and high throughput.
Также известна как
Compare GPT-5.4 Mini API pricing across 214 providers. Input prices range from $0.0002 to $7492.50 per million input. XShuLab Sub2API offers the lowest rate at $0.0002/M. 3 providers offer free API credits or a free tier.
| Провайдер | Вариант модели | Аудит | Входные данные ($/M) | Выходные данные ($/M) | Скорость (т/с) | Первый токен |
|---|---|---|---|---|---|---|
TokenClub API100% | gpt-5.4-mini | — | $0.750 | $4.50 | 245.0 t/s+48% | 2.62 s-34% |
Mentoe API99.9% | gpt-5.4-mini | — | $75.00 | $450.00 | 181.3 t/s+10% | 3.27 s-17% |
| gpt-5.4-mini | — | $0.100-87% | $0.600-87% | 142.5 t/s | 4.92 s | |
XShuLab Sub2API99.3% | gpt-5.4-mini | — | $0.0002-100% | $0.0012-100% | 139.4 t/s | 2.84 s-28% |
Mars HK0% | gpt-5.4-mini | — | $75.00 | $450.00 | 128.3 t/s | 4.45 s |
DNSHE0% | gpt-5.4-mini | — | Бесплатно | Бесплатно | 123.4 t/s | 8.53 s |
MyWebUI API84.3% | gpt-5.4-mini | — | $0.750 | $4.50 | 97.9 t/s | 4.04 s |
Tokeness99.9% | gpt-5.4-mini | — | $0.487-35% | $2.92-35% | 94.8 t/s | 2.34 s-41% |
柏拉图AI100% | gpt-5.4-mini | — | $0.103-86% | $0.616-86% | — | — |
| gpt-5.4-mini-2026-03-17 | — | $0.103-86% | $0.616-86% | — | — | |
V-API100% | gpt-5.4-mini | — | $0.750 | $4.50 | — | — |
| gpt-5.4-mini-2026-03-17 | — | $0.750 | $4.50 | — | — | |
| gpt-5.4-mini-high | — | $0.750 | $4.50 | — | — | |
| gpt-5.4-mini-low | — | $0.750 | $4.50 | — | — | |
| gpt-5.4-mini-medium | — | $0.750 | $4.50 | — | — | |
| gpt-5.4-mini-xhigh | — | $0.750 | $4.50 | — | — | |
老张API100% | gpt-5.4-mini | — | $0.719-4% | $4.32-4% | — | — |
PackyAPI100% | gpt-5.4-mini | — | $0.107-86% | $0.643-86% | — | — |
Cuz AI100% | gpt-5.4-mini | — | $0.750 | $4.50 | — | — |
AiXiaobai API100% | gpt-5.4-mini | — | $0.175-77% | $1.05-77% | — | — |
Данные о ценах из публичных API провайдеров
Сравните производительность по скорости и задержке у всех API-провайдеров.
| Провайдер | Скорость | Задержка | Тесты |
|---|---|---|---|
TokenClub API gpt-5.4-mini | 244.95 tok/s | 2.62s | 10 |
CPA/gpt-5.4-mini | 223.30 tok/s | 3.41s | 5 |
gpt-5.4-mini | 200.70 tok/s | 1.44s | 5 |
openai/gpt-5.4-mini | 195.10 tok/s | 1.10s | 5 |
gpt-5.4-mini | 181.26 tok/s | 3.27s | 10 |
gpt-5.4-mini | 154.80 tok/s | 4.49s | 5 |
gpt-5.4-mini | 144.75 tok/s | 5.78s | 5 |
gpt-5.4-mini | 142.47 tok/s | 4.92s | 10 |
gpt-5.4-mini | 139.42 tok/s | 2.84s | 25 |
gpt-5.4-mini | 128.34 tok/s | 4.45s | 15 |
gpt-5.4-mini | 123.38 tok/s | 8.53s | 5 |
gpt-5.4-mini | 112.29 tok/s | 0.88s | 5 |
gpt-5.4-mini | 107.95 tok/s | 1.95s | 15 |
gpt-5.4-mini | 97.89 tok/s | 4.04s | 5 |
gpt-5.4-mini | 94.81 tok/s | 2.34s | 5 |
gpt-5.4-mini | 92.94 tok/s | 4.06s | 5 |
gpt-5.4-mini | 50.48 tok/s | 7.77s | 10 |
Показано 1–17 из 17 провайдеров
gpt-5-4
OpenAI GPT-5.4 extends the GPT-5 family with stronger instruction following, deeper tool use, and improved performance on coding, math, and long-document analysis.
gpt-5-3-codex
OpenAI GPT-5.3 Codex is a code-specialized variant in the GPT-5 series, optimized for code generation, debugging, and software development tasks.
gpt-5-2
OpenAI GPT-5.2 is a GPT-5 series model emphasizing advanced reasoning, multimodal understanding, and high-quality outputs for complex enterprise workloads.
claude-opus-4-6
Anthropic Claude Opus 4.6 is the most capable Claude Opus tier, optimized for complex analysis, long-horizon coding, and high-stakes enterprise reasoning workloads.
claude-sonnet-4-6
Anthropic Claude Sonnet 4.6 extends the Sonnet line with improved tool use, coding reliability, and long-context performance for everyday production workloads.
claude-haiku-4-5
Anthropic Claude Haiku 4.5 delivers fast, low-cost responses while retaining solid instruction following for chat, classification, and lightweight coding.
Рейтинги основаны на тестах, предоставленных сообществом, и периодических зондах работоспособности. Носит рекомендательный характер, не является официальными данными.Artificial Analysis
