After a period of time, MaaS models are deprecated and typically replaced with newer model versions. To provide you with time to test and migrate to newer models, this page lists all models that are deprecated along with their shutdown date.
Anthropic's Claude 3 Sonnet
Anthropic's Claude 3 Sonnet is deprecated as of January 21, 2025 and will be shutdown on July 21, 2025. Claude 3 Sonnet is available to existing customers only.
Claude 3 Sonnet is engineered to be dependable for scaled AI deployments across a variety of use cases. Claude 3 Sonnet is optimized for the following use cases:
Data processing, including retrieval-augmented generation (RAG) and search retrieval.
Sales tasks, such as product recommendations, forecasting, and targeted marketing.
Time-saving tasks, such as code generation, quality control, and optical character recognition (OCR) in images.
Vision tasks, such as processing images to return text output. Also, analysis of charts, graphs, technical diagrams, reports, and other visual content.
The following table shows the maximum quotas and supported context length for Claude 3 Sonnet in each region.
| Region | Quotas | Supported context length |
|---|---|---|
us-east5 (Ohio) |
Up to 10 QPM, 30,000 TPM | 200,000 tokens |
Using Claude 3 Sonnet
For SDK and curl commands, use claude-3-sonnet@20240229 as the model name.
Pricing
For existing Anthropic's Claude 3 Sonnet users, pricing remains the same. For details, see the Pricing page.
Mistral AI models
Mistral Large (24.07) and Codestral (24.05) are deprecated as of January 30, 2025 and will be shutdown on July 29, 2025. These two models are available to existing customers only.
Mistral Large (24.07) (mistral-large@2407) is
Mistral AI's flagship model for text generation. It reaches top-tier
reasoning capabilities and can be used for complex multilingual tasks, including
text understanding, transformation, and code generation.
Codestral (24.05) (codestral@2405) is a generative model that
has been specifically designed and optimized for code generation tasks,
including fill-in-the-middle and code completion. Codestral (24.05) was trained
on more than 80 programming languages, enabling it to perform well on both
common and less common languages.
Supported regions
Mistral Large (24.07)
| Region | Quotas | Supported context length |
|---|---|---|
us-central1 |
Up to 60 QPM, 400,000 TPM | 128,000 tokens |
europe-west4 |
Up to 60 QPM, 400,000 TPM | 128,000 tokens |
Codestral (24.05)
| Region | Quotas | Supported context length |
|---|---|---|
us-central1 |
Up to 60 QPM, 400,000 TPM | 32,000 tokens |
europe-west4 |
Up to 60 QPM, 400,000 TPM | 32,000 tokens |
Pricing
For existing Mistral Large (24.07) and Codestral (24.05) users, pricing remains the same. For details, see the Pricing page.