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.
Claude 3.5 Sonnet v2
Claude 3.5 Sonnet v2 is deprecated as of August 20, 2025and will be shut down on February 19, 2026. Claude 3.5 Sonnet v2 is available to existing customers only.
Claude 3.5 Sonnet v2 is a state-of-the-art model for real-world software engineering tasks and agentic capabilities.
Try in Vertex AI View model card in Model Garden
claude-3-5-sonnet-v2@20241022
- Batch predictions Supported
- Prompt caching Supported
- Function calling Supported
- Extended thinking Not supported
- Count tokens Supported
- Limitation and specifications: See Vision in Anthropic's documentation
- Limitation and specifications: See PDF support in Anthropic's documentation
-
claude-3-5-sonnet-v2@20241022 - Launch stage: Generally available
- Release date: October 22, 2024
Model availability
(Includes fixed quota & Provisioned Throughput)
United States
-
us-east5
Europe
-
europe-west1
Global
-
global endpoint
ML processing
United States
-
Multi-region
Europe
-
Multi-region
us-east5:
- QPM: 90
- TPM: 540,000 (input and output)
- Context length: 200,000
europe-west1:
- QPM: 55
- TPM: 330,000 (input and output)
- Context length: 200,000
global endpoint:
- QPM: 25
- TPM: 140,000 (input and output)
- Context length: 200,000
Claude 3.5 Sonnet
Claude 3.5 Sonnet is deprecated as of August 20, 2025and will be shut down on February 19, 2026. Claude 3.5 Sonnet is available to existing customers only.
Claude 3.5 Sonnet outperforms Anthropic's Claude 3 Opus on a wide range of Anthropic's evaluations with the speed and cost of Anthropic's mid-tier model, Claude 3 Sonnet.
View model card in Model Garden
claude-3-5-sonnet@20240620
- Batch predictions Not supported
- Prompt caching Supported
- Function calling Supported
- Extended thinking Not supported
- Count tokens Supported
- Limitation and specifications: See Vision in Anthropic's documentation
- Limitation and specifications: See PDF support in Anthropic's documentation
-
claude-3-5-sonnet@20240620 - Launch stage: Generally available
- Release date: June 20, 2024
Model availability
(Includes fixed quota & Provisioned Throughput)
United States
-
us-east5
Europe
-
europe-west1
Asia pacific
-
asia-southeast1
ML processing
United States
-
Multi-region
Europe
-
Multi-region
Asia pacific
-
asia-southeast1
us-east5:
- QPM: 80
- TPM: 350,000 (input and output)
- Context length: 200,000
europe-west1:
- QPM: 130
- TPM: 600,000 (input and output)
- Context length: 200,000
asia-southeast1:
- QPM: 35
- TPM: 150,000 (input and output)
- Context length: 200,000
Jamba 1.5 Large
Jamba 1.5 Large is deprecated as of August 27, 2025and will be shut down on February 27, 2026. Jamba 1.5 Large is available to existing customers only.
AI21 Labs's Jamba 1.5 Large is well balanced across quality, throughput, and low cost.
View model card in Model Garden
jamba-1.5-large
-
jamba-1.5-large - Launch stage: Preview
- Release date: August 22, 2024
Model availability
United States
-
us-central1
Europe
-
europe-west4
ML processing
United States
-
Multi-region
us-central1:
- QPM: 20
- TPM: 20,000
- Context length: 256,000
europe-west4:
- QPM: 20
- TPM: 20,000
- Context length: 256,000
Jamba 1.5 Mini
Jamba 1.5 Mini is deprecated as of August 27, 2025and will be shut down on February 27, 2026. Jamba 1.5 Mini is available to existing customers only.
AI21 Labs's Jamba 1.5 Mini is well balanced across quality, throughput, and low cost.
View model card in Model Garden
jamba-1.5-mini
-
jamba-1.5-mini - Launch stage: Preview
- Release date: August 22, 2024
Model availability
United States
-
us-central1
Europe
-
europe-west4
ML processing
United States
-
Multi-region
us-central1:
- QPM: 50
- TPM: 60,000
- Context length: 256,000
europe-west4:
- QPM: 50
- TPM: 60,000
- Context length: 256,000
Mistral Nemo
Mistral Nemo is deprecated as of June 30, 2025and will be shut down on August 20, 2025. Mistral Nemo is available to existing customers only.
Mistral Nemo is Mistral AI's most cost efficient proprietary model. Use Mistral Nemo low-latency workloads and basic tasks that can be done in bulk, such as classification, customer support, and text generation.
View model card in Model Garden
mistral-nemo
-
mistral-nemo - Launch stage: Deprecated
- Release date: July 24, 2024
Model availability
United States
-
us-central1
Europe
-
europe-west4
ML processing
United States
-
Multi-region
Europe
-
Multi-region
us-central1:
- QPM: 60
- TPM: 400,000
- Context length: 128,000
europe-west4:
- QPM: 60
- TPM: 400,000
- Context length: 128,000
Claude 3 Opus
Anthropic's Claude 3 Opus is deprecated as of June 30, 2025and will be shut down on August 1, 2025. Claude 3 Opus is available to existing customers only.
Anthropic's Claude 3 Opus is a powerful AI model with top-level performance on highly complex tasks. It can navigate open-ended prompts and sight-unseen scenarios with remarkable fluency and human-like understanding. Claude 3 Opus is optimized for the following use cases:
-
Task automation, such as interactive coding and planning, or running complex actions across APIs and databases.
-
Research and development tasks, such as research review, brainstorming and hypothesis generation, and product testing.
-
Strategy tasks, such as advanced analysis of charts and graphs, financials and market trends, and forecasting.
-
Vision tasks, such as processing images to return text output. Also, analysis of charts, graphs, technical diagrams, reports, and other visual content.
View model card in Model Garden
claude-3-opus@20240229
- Batch predictions Not supported
- Prompt caching Supported
- Function calling Supported
- Extended thinking Not supported
- Count tokens Supported
- Limitation and specifications: See Vision in Anthropic's documentation
- Limitation and specifications: See PDF support in Anthropic's documentation
-
claude-3-opus@20240229 - Launch stage: Deprecated
- Release date: May 31, 2024
Model availability
(Includes fixed quota & Provisioned Throughput)
United States
-
us-east5
ML processing
United States
-
Multi-region
us-east5:
- QPM: 20
- TPM: 105,000 (input and output)
- Context length: 200,000
