Pricing
Text-to-Video
| Model | Input | Output |
|---|---|---|
| MaaS Haiper Video | - | US$0.06 / Second |
Speech Models
| Model | Input | Output |
|---|---|---|
| MaaS-Ele | US$0.3 / 1K Characters | - |
| MaaS-nar | - | US$0.2 / minute |
| MaaS ASpeech | US$0.015 / 1K Characters; HD：US$0.03 / 1K Characters |
- |
| MaaS OSpeech | US$0.015 / 1K Characters | - |
| MaaS OSpeech_HD | US$0.03 / 1K Characters | - |
| MaaS Whisper | US$0.36 / hour | - |
Text-to-Text
MaaS-Llama
| Model | Input | Output |
|---|---|---|
| MaaS Llama 3.1 8B Instruct | US$0.0002 / 1K tokens | US$0.0002 / 1K tokens |
| MaaS Llama 3.1 70B Instruct | US$0.0009 / 1K tokens | US$0.0009 / 1K tokens |
MaaS-Ge
The MaaS-Ge model exhibits exceptional precision, adeptness at multi-tasking, and remarkable scalability, facilitating the efficient execution of complex tasks across diverse application domains.
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS-1.5 Flash | =<128K | Image：US$0.0001315 / image Video：US$0.0001315 / second Text：US$0.000125 / 1K tokens Audio：US$0.0000125 / second |
Text：US$0.000375 / 1K tokens |
| MaaS-1.5 Flash | >128K | Image：US$0.000263 / image Video：US$0.000263 / second Text：US$0.00025 / 1K tokens Audio：US$0.000025 / second |
Text：US$0.00075 / 1K tokens |
| MaaS-1.5 Pro | =<128K | Image：US$0.001315 / image Video：US$0.001315 / second Text：US$0.00125 / 1K tokens Audio：US$0.000125 / second |
Text：US$0.00375 / 1K tokens |
| MaaS-1.5 Pro | >128K | Image：US$0.00263 / image Video：US$0.00263 / second Text：US$0.0025 / 1K tokens Audio：US$0.00025 / second |
Text：US$0.0075 / 1K tokens |
| MaaS-1.0 Pro | =<128K | Image：US$0.0025 / image Video：US$0.002 / second Text：US$0.000125 / 1K tokens |
Text：US$0.000375 / 1K tokens |
*All quoted prices are exclusive of taxes.
MaaS-C
MaaS-C is an exemplary natural language processing model renowned for its remarkable prowess in comprehension and generation. Its unparalleled ability lies in grasping intricate semantic relationships and contextual information, crafting text that is of superior quality and flows effortlessly, akin to the beauty of nature.
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS-3.5 Sonnet | 200K | US$0.003 / 1K tokens | US$0.015 / 1K tokens |
| MaaS-3 Sonnet | 200K | US$0.003 / 1K tokens | US$0.015 / 1K tokens |
| MaaS-3 Haiku | 200K | US$0.00025 / 1K tokens | US$0.00125 / 1K tokens |
*All quoted prices are exclusive of taxes.
MaaS-GP
With 128k context, fresher knowledge and the broadest set of capabilities, MaaS-4o is more powerful than MaaS-4 and offered at a lower price.
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS o1 preview 2024-09-12 | 128K | US$0.015 / 1K tokens | US$0.060 / 1K tokens |
| MaaS o1 mini 2024-09-12 | 128K | US$0.003 / 1K tokens | US$0.012 / 1K tokens |
| MaaS 4o 2024-08-06 | 128K | US$0.0025 / 1K tokens | US$0.01 / 1K tokens |
| MaaS-4o mini | 128K | US$0.000150 / 1K tokens | US$0.000600 / 1K tokens |
| MaaS-4o | 128K | US$0.005 / 1K tokens | US$0.015 / 1K tokens |
*All quoted prices are exclusive of taxes.
With broad general knowledge and domain expertise, MaaS--4 can follow complex instructions in natural language and solve difficult problems with accuracy. The pricing for MaaS-4 is as follows:
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS-4 | 8K | US$0.03 / 1K tokens | US$0.06 / 1K tokens |
| MaaS-4-32K | 32K | US$0.06 / 1K tokens | US$0.12 / 1K tokens |
*All quoted prices are exclusive of taxes.
MaaS-4 Turbo is a variant of the MaaS-4 model optimized for faster response times and higher efficiency, making it more suitable for applications requiring real-time interactions and lower computational costs.
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS-4-turbo | 128K | US$0.01 / 1K tokens | US$0.03 / 1K tokens |
*All quoted prices are exclusive of taxes.
MaaS-3.5 Turbo models are capable and cost-effective.
MaaS-3.5-turbo-0125 is the flagship model of this family, supports a 16K context window and is optimized for dialog.
MaaS-3.5-turbo-instruct is an Instruct model and only supports a 4K context window.
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS-3.5-turbo-0125 | 16K | US$0.0005 / 1K tokens | US$0.0015 / 1K tokens |
| MaaS-3.5-turbo-instruct | 4K | US$0.0015 / 1K tokens | US$0.0020 / 1K tokens |
*All quoted prices are exclusive of taxes.
Embedding
Build advanced search, clustering, topic modeling, and classification functionality with our embeddings offering.
| Model | Usage |
|---|---|
| MaaS-embedding-3-small | US$0.00002 / 1K tokens |
| MaaS-embedding-3-large | US$0.00013 / 1K tokens |
| MaaS-Ada v2 | US$0.00010 / 1K tokens |
*All quoted prices are exclusive of taxes.
Text-to-Image
Build MaaS-DALL·E directly into your apps to generate and edit novel images and art. MaaS-DALL·E 3 is the highest quality model and MaaS-DALL·E 2 is optimized for lower cost.
| Model | Quality | Resolution | Price |
|---|---|---|---|
| MaaS-DALL·E 3 | Standard | 1024×1024 | $0.040 / image |
| MaaS-DALL·E 3 | Standard | 1024×1792, 1792×1024 | $0.080 / image |
| MaaS-DALL·E 3 | HD | 1024×1024 | $0.080 / image |
| MaaS-DALL·E 3 | HD | 1024×1792, 1792×1024 | $0.120 / image |
| MaaS-DALL·E 2 | - | 1024×1024 | $0.020 / image |
| MaaS-DALL·E 2 | - | 512×512 | $0.018 / image |
| MaaS-DALL·E 2 | - | 256×256 | $0.016 / image |
*All quoted prices are exclusive of taxes.
Older Models
We continue to improve our models and periodically retire older, less used models.
| Model | Context | Input | Output |
|---|---|---|---|
| MaaS-4-0125-preview | 128K | US$0.0100 / 1K tokens | US$0.0300 / 1K tokens |
| MaaS-4-1106-preview | 128K | US$0.0100 / 1K tokens | US$0.0300 / 1K tokens |
| MaaS-4-vision-preview | 128K | US$0.0100 / 1K tokens | US$0.0300 / 1K tokens |
| MaaS-3.5-turbo-1106 | 16K | US$0.0010 / 1K tokens | US$0.0020 / 1K tokens |
| MaaS-3.5-turbo-0613 | 4K | US$0.0015 / 1K tokens | US$0.0020 / 1K tokens |
| MaaS-3.5-turbo-16k-0613 | 16K | US$0.0030 / 1K tokens | US$0.0040 / 1K tokens |
| MaaS-3.5-turbo-0301 | 4K | US$0.0020 / 1K tokens | US$0.0020 / 1K tokens |
*All quoted prices are exclusive of taxes.