TeleSpeechASR is available through 8 API providers on LMSpeed. Compare API pricing from $0.0071 to $535.71 per million input tokens across providers.
TeleSpeechASR is a speech-to-text model, designed for accurate audio transcription and recognition.
Также известна как
Compare TeleSpeechASR API pricing across 8 providers. Input prices range from $0.0071 to $535.71 per million input. 小水管 API offers the lowest rate at $0.0071/M.
| Провайдер | Вариант модели | Аудит | Входные данные ($/M) | Выходные данные ($/M) | Скорость (т/с) | Первый токен |
|---|---|---|---|---|---|---|
小水管 API100% | telespeechasr | — | $0.0071 | $0.0071 | — | — |
天絮 API100% | TeleAI/TeleSpeechASR | — | $535.71 | $535.71 | — | — |
素墨API100% | telespeechasr | — | $0.010 | $0.010 | — | — |
钱多多 API100% | TeleAI/TeleSpeechASR | — | $0.979 | $0.979 | — | — |
SWT-API98.3% | telespeechasr | — | $0.037 | $0.037 | — | — |
Koyeb AI Gateway98.3% | telespeechasr | — | $0.037 | $0.037 | — | — |
ChatST API99.9% | TeleSpeechASR | — | $6.99 | $6.99 | — | — |
WAADRI0% | TeleAI/TeleSpeechASR | — | $2.00 | $2.00 | — | — |
Данные о ценах из публичных API провайдеров
kimi-k2-5
Moonshot Kimi K2.5 is an open-weight multimodal agent model with native vision and text input, strong coding performance, and a 256K context window.
glm-5
Zhipu GLM-5 is Zhipu flagship GLM series model with enhanced reasoning, agent capabilities, and strong performance on Chinese enterprise and coding scenarios.
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.
glm-4-5-air
Zhipu AI GLM-4.5 Air is a fast and efficient language model in the GLM series, optimized for quick responses and high throughput.
deepseek-v3-2
DeepSeek V3.2 is an upgraded V3-series MoE model with stronger reasoning, coding, and math performance, widely available through OpenAI-compatible API relays.
deepseek-v4-flash
DeepSeek V4 Flash is a fast, cost-efficient language model in the DeepSeek V4 family, optimized for low-latency chat, coding assistance, and high-throughput API workloads while retaining strong reasoning quality.
Рейтинги основаны на тестах, предоставленных сообществом, и периодических зондах работоспособности. Носит рекомендательный характер, не является официальными данными.
