已收录报价的服务商
75 家推理服务商目前对 269 个追踪模型有端点报价,均经 OpenRouter 观察,另有 DeepSeek、Z.AI、Groq、Xiaomi 的有来源复核的直连报价。这是「谁的价格在这里」的清单,不是所有 AI 厂商的目录。
已覆盖
- 842 条有效端点报价(含不可用共 842 条,保留供参考)。
- 成对输入/输出费率、已公布的缓存读/写价、量化精度、上下文和声明参数。
- 直连报价币种:CNY、USD。
- 价格截至 2026-11-22 · 自 2026-09-29 起追踪 · 共 148 次检查。
尚未覆盖
- 大多数服务商的官方直连价格页;下方直连报价是目前的例外。
- 直连报价之外的区域定价、上下文阶梯和用量折扣。
- embedding、语音、图像和视频计费;目前只比较文本生成。
- 免费额度、数据使用条款和并发限制。
- 实测延迟、成功率和质量;价格是唯一排序维度。
经 OpenRouter 的端点服务商
| 服务商 | 模型 | 端点(可用 / 总数) | 已公布缓存价 | 最近价格变化 | 操作 |
|---|---|---|---|---|---|
| AionLabs | 6 | 6 / 6 | 5 / 6 | 自 2026-09-29 起无变化 | 比较 |
| AkashML | 1 | 1 / 1 | 1 / 1 | 比较 | |
| Alibaba | 22 | 29 / 29 | 22 / 29 | 比较 | |
| Amazon Bedrock | 20 | 28 / 28 | 5 / 28 | 自 2026-09-29 起无变化 | 比较 |
| Ambient | 2 | 2 / 2 | 2 / 2 | 比较 | |
| Ant Digital | 1 | 1 / 1 | 0 / 1 | 自 2026-11-15 起无变化 | 比较 |
| Azure | 11 | 11 / 11 | 4 / 11 | 自 2026-09-29 起无变化 | 比较 |
| Baidu | 7 | 7 / 7 | 6 / 7 | 比较 | |
| BaseTen | 13 | 23 / 23 | 23 / 23 | 比较 | |
| Cerebras | 2 | 2 / 2 | 2 / 2 | 自 2026-09-29 起无变化 | 比较 |
| Chutes | 7 | 7 / 7 | 7 / 7 | 比较 | |
| Cloudflare | 10 | 10 / 10 | 6 / 10 | 比较 | |
| Cohere | 11 | 11 / 11 | 7 / 11 | 比较 | |
| CoreWeave | 21 | 21 / 21 | 21 / 21 | 比较 | |
| Crusoe | 7 | 7 / 7 | 7 / 7 | 比较 | |
| Darkbloom | 12 | 12 / 12 | 12 / 12 | 比较 | |
| Decart | 1 | 1 / 1 | 1 / 1 | 比较 | |
| DeepInfra | 81 | 134 / 134 | 56 / 134 | 比较 | |
| DekaLLM | 1 | 1 / 1 | 0 / 1 | 自 2026-09-29 起无变化 | 比较 |
| DigitalOcean | 16 | 16 / 16 | 16 / 16 | 比较 | |
| Doubleword | 1 | 1 / 1 | 1 / 1 | 自 2026-11-20 起无变化 | 比较 |
| Featherless | 2 | 2 / 2 | 2 / 2 | 比较 | |
| Fireworks | 13 | 21 / 21 | 21 / 21 | 比较 | |
| Friendli | 3 | 3 / 3 | 2 / 3 | 自 2026-09-29 起无变化 | 比较 |
| 5 | 5 / 5 | 1 / 5 | 自 2026-09-29 起无变化 | 比较 | |
| Google AI Studio | 3 | 3 / 3 | 0 / 3 | 自 2026-09-29 起无变化 | 比较 |
| Groq | 4 | 4 / 4 | 4 / 4 | 自 2026-09-29 起无变化 | 比较 |
| Inceptron | 6 | 6 / 6 | 6 / 6 | 比较 | |
| Inferact vLLM | 1 | 1 / 1 | 1 / 1 | 比较 | |
| InferenceNet | 9 | 22 / 22 | 22 / 22 | 比较 | |
| Io Net | 4 | 4 / 4 | 4 / 4 | 比较 | |
| Ionstream | 3 | 3 / 3 | 3 / 3 | 比较 | |
| KT | 7 | 7 / 7 | 7 / 7 | 比较 | |
| Liquid | 3 | 3 / 3 | 2 / 3 | 自 2026-11-21 起无变化 | 比较 |
| Makora | 9 | 9 / 9 | 9 / 9 | 比较 | |
| Mancer 2 | 10 | 10 / 10 | 10 / 10 | 比较 | |
| Mistral | 23 | 36 / 36 | 31 / 36 | 比较 | |
| Modal | 10 | 10 / 10 | 10 / 10 | 自 2026-09-29 起无变化 | 比较 |
| ModelRun | 7 | 7 / 7 | 7 / 7 | 比较 | |
| Moonshot AI | 4 | 6 / 6 | 6 / 6 | 自 2026-10-24 起无变化 | 比较 |
| Morph | 14 | 15 / 15 | 11 / 15 | 比较 | |
| Near AI | 4 | 4 / 4 | 4 / 4 | 比较 | |
| Nebius | 20 | 20 / 20 | 19 / 20 | 比较 | |
| Nex AGI | 2 | 2 / 2 | 2 / 2 | 自 2026-09-29 起无变化 | 比较 |
| NextBit | 5 | 5 / 5 | 3 / 5 | 比较 | |
| Novita | 19 | 19 / 19 | 12 / 19 | 比较 | |
| Nvidia | 7 | 7 / 7 | 0 / 7 | 自 2026-09-29 起无变化 | 比较 |
| OpenAI | 8 | 8 / 8 | 5 / 8 | 自 2026-09-29 起无变化 | 比较 |
| OpenInference | 3 | 3 / 3 | 3 / 3 | 比较 | |
| Parasail | 39 | 46 / 46 | 39 / 46 | 比较 | |
| Particle AI | 4 | 7 / 7 | 6 / 7 | 比较 | |
| Pearl Research | 1 | 1 / 1 | 1 / 1 | 自 2026-11-19 起无变化 | 比较 |
| Perceptron | 2 | 2 / 2 | 0 / 2 | 自 2026-09-29 起无变化 | 比较 |
| Perplexity | 2 | 2 / 2 | 0 / 2 | 自 2026-09-29 起无变化 | 比较 |
| PrimaLabs | 4 | 4 / 4 | 4 / 4 | 比较 | |
| PrimeIntellect | 14 | 14 / 14 | 12 / 14 | 比较 | |
| Recursive | 1 | 1 / 1 | 0 / 1 | 自 2026-10-22 起无变化 | 比较 |
| Reka | 6 | 6 / 6 | 4 / 6 | 比较 | |
| Relace | 13 | 13 / 13 | 11 / 13 | 比较 | |
| River | 5 | 5 / 5 | 5 / 5 | 比较 | |
| Sail Research | 10 | 15 / 15 | 15 / 15 | 比较 | |
| SambaNova | 1 | 1 / 1 | 0 / 1 | 自 2026-09-29 起无变化 | 比较 |
| Seed | 1 | 1 / 1 | 0 / 1 | 自 2026-09-29 起无变化 | 比较 |
| SiliconFlow | 37 | 37 / 37 | 22 / 37 | 比较 | |
| StepFun | 1 | 1 / 1 | 1 / 1 | 自 2026-09-29 起无变化 | 比较 |
| StreamLake | 1 | 1 / 1 | 0 / 1 | 自 2026-09-29 起无变化 | 比较 |
| Tencent | 3 | 3 / 3 | 0 / 3 | 自 2026-09-29 起无变化 | 比较 |
| Tenstorrent | 5 | 5 / 5 | 5 / 5 | 比较 | |
| Together | 24 | 24 / 24 | 22 / 24 | 比较 | |
| Unbiased | 3 | 3 / 3 | 3 / 3 | 自 2026-10-01 起无变化 | 比较 |
| Venice | 30 | 32 / 32 | 26 / 32 | 比较 | |
| VoyageAI by MongoDB | 5 | 5 / 5 | 0 / 5 | 自 2026-11-04 起无变化 | 比较 |
| Wafer | 10 | 19 / 19 | 19 / 19 | 比较 | |
| Xiaomi | 3 | 3 / 3 | 3 / 3 | 自 2026-09-29 起无变化 | 比较 |
| Z.AI | 15 | 15 / 15 | 15 / 15 | 比较 |
有明确来源复核的直连报价
按官方价格页读取,币种为服务商实际计费币种,最多7天内复核,来源或决策可能提前到期。这是直连覆盖的起点,不是全部。
DeepSeek · DeepSeek-V4.1-Flash-1017
标准 · CNY · 地区未注明 · 核验于 2026-11-20
输入 ¥2.00 / 输出 ¥8.00
官方定价计费条件
Official current model version: DeepSeek-V4.1-Flash-1017; API identifier: deepseek-flash. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: peak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: CNY. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in peak window. Cache miss. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Includes thinking tokens. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Cache hit. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4.1-Flash-1017
谷时 · CNY · 地区未注明 · 核验于 2026-11-20
输入 ¥1.00 / 输出 ¥4.00
官方定价计费条件
Official current model version: DeepSeek-V4.1-Flash-1017; API identifier: deepseek-flash. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: offpeak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: CNY. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in offpeak window. Cache miss. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Includes thinking tokens. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Cache hit. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4-Pro-0813
标准 · CNY · 地区未注明 · 核验于 2026-11-20
输入 ¥9.00 / 输出 ¥27.00
官方定价计费条件
Official current model version: DeepSeek-V4-Pro-0813; API identifier: deepseek-v4-pro. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: peak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: CNY. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in peak window. Cache miss. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Includes thinking tokens. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Cache hit. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4-Pro-0813
谷时 · CNY · 地区未注明 · 核验于 2026-11-20
输入 ¥4.50 / 输出 ¥13.50
官方定价计费条件
Official current model version: DeepSeek-V4-Pro-0813; API identifier: deepseek-v4-pro. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: offpeak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: CNY. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in offpeak window. Cache miss. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Includes thinking tokens. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Cache hit. Peak: Monday–Friday 09:00–12:00 and 14:00–18:00 Beijing time, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4.1-Flash-1017
标准 · USD · 地区未注明 · 核验于 2026-11-20
输入 US$0.30 / 输出 US$1.20
官方定价计费条件
Official current model version: DeepSeek-V4.1-Flash-1017; API identifier: deepseek-flash. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: peak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: USD. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in peak window. Cache miss. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Includes thinking tokens. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Cache hit. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4.1-Flash-1017
谷时 · USD · 地区未注明 · 核验于 2026-11-20
输入 US$0.15 / 输出 US$0.60
官方定价计费条件
Official current model version: DeepSeek-V4.1-Flash-1017; API identifier: deepseek-flash. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: offpeak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: USD. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in offpeak window. Cache miss. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Includes thinking tokens. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Cache hit. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4-Pro-0813
标准 · USD · 地区未注明 · 核验于 2026-11-20
输入 US$1.32 / 输出 US$3.96
官方定价计费条件
Official current model version: DeepSeek-V4-Pro-0813; API identifier: deepseek-v4-pro. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: peak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: USD. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in peak window. Cache miss. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Includes thinking tokens. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in peak window. Cache hit. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
DeepSeek · DeepSeek-V4-Pro-0813
谷时 · USD · 地区未注明 · 核验于 2026-11-20
输入 US$0.66 / 输出 US$1.98
官方定价计费条件
Official current model version: DeepSeek-V4-Pro-0813; API identifier: deepseek-v4-pro. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays. Selected billing window: offpeak. Cache-hit input uses the cache-read rate; cache-miss input uses the input rate. Thinking tokens are billed as output. Context: 1M tokens; documented maximum output: 384K. Standalone Search and Fetch APIs are billed per successful request. Since November 18, 2026 at 12:00 Beijing time, model-triggered server-side search and fetch carry no separate tool fee; returned content is billed as model token usage. Token reference subtotals exclude independent API calls. Billing currency: USD. Deployment region and account availability are not established by this price page; region is unspecified, not inferred from currency or document language. Supports documented JSON output and tool calls, with thinking enabled by default and thinking/reasoning_effort controls. Control documentation: https://api-docs.deepseek.com/guides/thinking_mode/; internal evidence SHA-256: 5ab9a7b03778a41ea37046b88936a3066fcfe9085d6b7436b5fc0ec371b08a88. Output quality and strict schema adherence are not established.
Applies in offpeak window. Cache miss. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Includes thinking tokens. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Applies in offpeak window. Cache hit. Peak: Monday–Friday 01:00–04:00 and 06:00–10:00 UTC, excluding Chinese public holidays. Off-peak: all other times, including weekends and Chinese public holidays.
Z.AI · GLM-5.3-Flash
标准 · CNY · 地区未注明 · 核验于 2026-11-20
输入 ¥0.80 / 输出 ¥2.80
官方定价计费条件
Standard token rates. Cache storage listed as limited-time free; no permanent zero rate recorded.
Z.AI · GLM-5.3-Flash
标准 · USD · 地区未注明 · 核验于 2026-11-20
输入 US$0.15 / 输出 US$0.50
官方定价计费条件
Standard token rates. Cache storage listed as limited-time free; no permanent zero rate recorded.
Groq · Whisper Large V3 Turbo
标准 · USD · 地区未注明 · 核验于 2026-11-18
duration: 0.04 USD / hour
每文件最低计费: 10 秒
官方定价计费条件
Standard synchronous transcription, production listing; USD 0.04 per audio hour; minimum billed duration 10 seconds per request; minimum accepted file 0.01 seconds. Account/file-size limits, taxes, custom contracts, Batch and other services excluded; no quality or access guarantee. Currency proof: c5b3de857f41a23325c2dd256737dc5d1ba93af12c6b387975b079acecd9def0 (full_document_email_links_v1); raw response e3d388b5f00f79e2c4a50e62829af8f3f006c567151d20856bef1160ebca55ba. Source model listing is production; not tested account availability. Unit-specific known charges only, not a complete bill.
Groq · Whisper Large V3
标准 · USD · 地区未注明 · 核验于 2026-11-18
duration: 0.111 USD / hour
每文件最低计费: 10 秒
官方定价计费条件
Standard synchronous transcription, production listing; USD 0.111 per audio hour; minimum billed duration 10 seconds per request; minimum accepted file 0.01 seconds. Account/file-size limits, taxes, custom contracts, Batch and other services excluded; no quality or access guarantee. Currency proof: c5b3de857f41a23325c2dd256737dc5d1ba93af12c6b387975b079acecd9def0 (full_document_email_links_v1); raw response e3d388b5f00f79e2c4a50e62829af8f3f006c567151d20856bef1160ebca55ba. Source model listing is production; not tested account availability. Unit-specific known charges only, not a complete bill.
Xiaomi · MiMo-V2.6-Flash
标准 · USD · 地区未注明 · 核验于 2026-11-20
输入 US$0.14 / 输出 US$0.28
官方定价计费条件
Pay-as-you-go token rates for the stated currency. Deployment region and account eligibility unspecified.
Xiaomi · MiMo-V2.6-Flash
标准 · CNY · 地区未注明 · 核验于 2026-11-20
输入 ¥1.00 / 输出 ¥2.00
官方定价计费条件
Pay-as-you-go token rates for the stated currency. Deployment region and account eligibility unspecified.
模型开发商和其他 AI 公司另见 厂商目录;出现在那里不代表其价格已在此追踪。