qwen3-instruct
Разработчик: Alibaba
Alibaba Qwen3 Instruct is an instruction-tuned variant in the Qwen3 family, optimized for chat, tool use, and general assistant applications.
Также известна как
Qwen3 Instruct by Alibaba is available through 45 API providers on LMSpeed. Compare API pricing from $0.0010 to $75.00 per million input tokens across providers. Free API access is offered by 1 provider. In speed benchmarks, the fastest provider reaches 910 tok/s.
Compare Qwen3 Instruct API pricing across 44 providers. Input prices range from $0.0010 to $75.00 per million input. 10dian-API offers the lowest rate at $0.0010/M. 1 provider offers free API credits or a free tier.
| Провайдер | Вариант модели | Аудит | Входные данные ($/M) | Выходные данные ($/M) | Скорость (т/с) | Первый токен |
|---|---|---|---|---|---|---|
V-API100% | Qwen3-235B-A22B-Instruct-2507 | — | $0.040 | $0.040 | — | — |
| qwen3-30b-a3b-instruct-2507 | — | $0.750 | $3.00 | — | — | |
| qwen3-235b-a22b-instruct-2507 | — | $2.00 | $8.00 | — | — | |
柏拉图AI100% | qwen3-30b-a3b-instruct-2507 | — | $0.104 | $0.411 | — | — |
| qwen3-235b-a22b-instruct-2507 | — | $0.274 | $1.10 | — | — | |
天絮 API100% | Qwen/Qwen3-30B-A3B-Instruct-2507 | — | $0.714 | $2.14 | — | — |
uglycat100% | qwen-3-235b-a22b-instruct-2507 | — | $0.010 | $0.010 | — | — |
AAAI100% | Qwen3-235B-A22B-Instruct-2507 | — | $0.040 | $0.040 | — | — |
| qwen3-30b-a3b-instruct-2507 | — | $0.750 | $3.00 | — | — | |
| qwen3-235b-a22b-instruct-2507 | — | $2.00 | $8.00 | — | — | |
Seamee API100% | Qwen/Qwen3-235B-A22B-Instruct-2507 | — | $0.090 | $0.600 | — | — |
OpenRouter Fans100% | Qwen/Qwen3-30B-A3B-Instruct-2507 | — | $0.220 | $0.880 | — | — |
SMLC666 API100% | qwen-3-235b-a22b-instruct-2507 | — | $0.600 | $1.20 | — | — |
| qwen3-235b-a22b-instruct | — | $75.00 | $75.00 | — | — | |
| qwen3-235b-a22b-instruct-2507 | — | $0.750 | $2.25 | — | — | |
钱多多 API100% | qwen/qwen3-30b-a3b-instruct-2507 | — | $0.367 | $1.47 | — | — |
| qwen3-30b-a3b-instruct-2507 | — | $0.367 | $1.47 | — | — | |
| Qwen/Qwen3-235B-A22B-Instruct-2507 | — | $0.979 | $3.92 | — | — | |
| qwen3-235b-a22b-instruct-2507 | — | $0.979 | $3.92 | — | — | |
| Qwen/Qwen3-30B-A3B-Instruct-2507 | — | $36.71 | $36.71 | — | — |
Данные о ценах из публичных API провайдеров
Сравните производительность по скорости и задержке у всех API-провайдеров.
| Провайдер | Скорость | Задержка | Тесты |
|---|---|---|---|
GPT Load (PP.UA) qwen-3-235b-a22b-instruct-2507 | 910.44 tok/s | 0.50s | 5 |
qwen-3-235b-a22b-instruct-2507 | 910.42 tok/s | 0.63s | 5 |
qwen-3-235b-a22b-instruct-2507 | 572.89 tok/s | 2.84s | 70 |
accounts/fireworks/models/qwen3-235b-a22b-instruct-2507 | 78.64 tok/s | 0.62s | 5 |
qwen3-235b-a22b-instruct-2507 | 65.96 tok/s | 1.19s | 5 |
Qwen3-235B-A22B-Instruct-2507 | 43.21 tok/s | 1.05s | 5 |
qwen3-235b-a22b-instruct-2507 | 43.21 tok/s | 0.95s | 15 |
qwen3-235b-a22b-instruct | 24.36 tok/s | 0.39s | 10 |
Qwen/Qwen3-235B-A22B-Instruct-2507 | 20.96 tok/s | 1.49s | 15 |
Показано 1–9 из 9 провайдеров
qwen3
Alibaba Qwen3 is the Qwen family's flagship LLM series with dense and MoE variants, seamless thinking/non-thinking modes, and leading open-source performance in math, code, and agent tasks.
deepseek-v3-2
DeepSeek V3.2 is a large language model in the DeepSeek V3 series, offering advanced reasoning, code generation, and multimodal capabilities.
deepseek-v3
DeepSeek V3 is DeepSeek flagship MoE language model with 671B total parameters, delivering strong performance in reasoning, coding, and multilingual tasks at competitive inference cost.
gemini-2-5-flash
Google Gemini 2.5 Flash is a fast and efficient language model in the Gemini series, optimized for quick responses and high throughput.
gpt-oss
GPT-OSS is an open-source language model offering advanced reasoning, code generation, and multimodal capabilities.
deepseek-v3-1
DeepSeek V3.1 is an open-weights style frontier model from DeepSeek with strong math, coding, and Chinese-English bilingual reasoning.
Рейтинги основаны на тестах, предоставленных сообществом, и периодических зондах работоспособности. Носит рекомендательный характер, не является официальными данными.
