Скорость вывода
8453.3 tok/s
8751.7 tok/s
88Claude Opus 4.550.4 tok/s
8949.9 tok/s
Claude Opus 4.5 by Anthropic is available through 144 API providers on LMSpeed. Compare API pricing from $0.0014 to $49950.00 per million input tokens across providers. Free API access is offered by 3 providers. Context window: 200,000. In speed benchmarks, the fastest provider reaches 70 tok/s.
Anthropic Claude Opus 4.5 is the flagship Claude model for the hardest reasoning, research, and multi-step agent tasks where accuracy matters more than speed.
Также известна как
Compare Claude Opus 4.5 API pricing across 141 providers. Input prices range from $0.0014 to $49950.00 per million input. XShuLab Sub2API offers the lowest rate at $0.0014/M. 3 providers offer free API credits or a free tier.
| Провайдер | Вариант модели | Аудит | Входные данные ($/M) | Выходные данные ($/M) | Скорость (т/с) | Первый токен |
|---|---|---|---|---|---|---|
| claude-opus-4-5-20251101 | — | $0.685-86% | $3.42-86% | 45.9 t/s | 2.49 s | |
柏拉图AI100% | claude-opus-4-5-20251101 | — | $0.685-86% | $3.42-86% | — | — |
V-API100% | claude-opus-4-5-20251101 | — | $5.00 | $25.00 | — | — |
老张API100% | claude-opus-4-5-20251101 | — | $4.79-4% | $23.97-4% | — | — |
PackyAPI100% | claude-opus-4-5-20251101 | — | $0.714-86% | $3.57-86% | — | — |
小水管 API100% | claude-opus-4-5-20251101 | — | Бесплатно | Бесплатно | — | — |
MapleLeaf API100% | claude-opus-4.5 | — | Бесплатно | Бесплатно | — | — |
CodeXE100% | claude-opus-4.5 | — | $0.050-99% | $0.050-100% | — | — |
AiXiaobai API100% | claude-opus-4-5-20251101 | — | $1.16-77% | $5.82-77% | — | — |
速创API99.9% | claude-opus-4-5-20251101 | — | $5.00 | $25.00 | — | — |
VSLLM99.7% | claude-opus-4-5-20251101 | 7210086100 | $0.182-96% | $0.182-99% | — | — |
ApiToken Online82.3% | claude-opus-4-5-20251101 | — | $121.67 | $608.33 | — | — |
Seamee API100% | claude-opus-4.5 | — | Бесплатно | Бесплатно | — | — |
OpenRouter Fans100% | claude-opus-4-5-20251101 | — | $0.710-86% | $3.55-86% | — | — |
GuaiHub100% | claude-opus-4-5-20251101 | — | $0.616-88% | $3.08-88% | — | — |
FineOneAPI100% | claude-opus-4-5-20251101 | — | $5.00 | $25.00 | — | — |
AAAI100% | claude-opus-4-5-20251101 | — | $5.00 | $25.00 | — | — |
UnifyLLM100% | claude-opus-4-5-20251101 | — | $5.00 | $25.00 | — | — |
毫秒API100% | claude-opus-4-5-20251101 | — | $5.00 | $25.00 | — | — |
Newagiai100% | claude-opus-4-5-20251101 | — | $0.685-86% | $3.42-86% | — | — |
Данные о ценах из публичных API провайдеров
Сравните производительность по скорости и задержке у всех API-провайдеров.
| Провайдер | Скорость | Задержка | Тесты |
|---|---|---|---|
Rnglg2 API claude-opus-4.5 | 69.66 tok/s | 4.08s | 5 |
claude-opus-4-5-20251101 | 59.42 tok/s | 2.79s | 5 |
claude-opus-4-5-20251101 | 56.48 tok/s | 2.60s | 15 |
claude-opus-4-5 | 53.14 tok/s | 2.30s | 5 |
claude-opus-4-5-20251101 | 52.45 tok/s | 5.33s | 5 |
claude-opus-4-5-20251101 | 45.89 tok/s | 2.49s | 5 |
claude-opus-4-5-20251101 | 43.11 tok/s | 3.21s | 5 |
claude-opus-4-5-20251101 | 42.91 tok/s | 3.20s | 5 |
Показано 1–8 из 8 провайдеров
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-haiku-4-5
Anthropic Claude Haiku 4.5 delivers fast, low-cost responses while retaining solid instruction following for chat, classification, and lightweight coding.
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-sonnet-4-5
Anthropic Claude Sonnet 4.5 balances strong reasoning, coding, and agent capabilities with lower latency and cost than Opus, making it a practical default for production assistants.
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-5-2
OpenAI GPT-5.2 is a GPT-5 series model emphasizing advanced reasoning, multimodal understanding, and high-quality outputs for complex enterprise workloads.
Рейтинги основаны на тестах, предоставленных сообществом, и периодических зондах работоспособности. Носит рекомендательный характер, не является официальными данными.Artificial Analysis
