AI Model Support in Appian

This page tracks the lifecycle status and support for generative AI (artificial intelligence) models used by Appian. It provides information about supported models, regional availability, deprecation timelines, and guidance for administrators managing AI model configurations.

As AI models are added, updated, or deprecated, refer to this page to:

• Understand which AI models are supported.
• Plan for model deprecations and transitions.
• Identify regional restrictions that may affect feature availability.
• Take required administrative actions to maintain AI functionality.

Use the interactive tool below to explore which models power each feature, view their processing regions, and understand which regions your data may be processed in:

Supported AI models in Appian

The following Appian-provided AI models are supported in Appian:

Model	Availability	Notes
Amazon Nova 2 Lite	Requires processing region	Fast model for text and vision tasks; not available for GovCloud
Amazon Textract	Multiple regions	Document analysis and text extraction
Claude Haiku 3.5	Limited regions	Fast, efficient language model
Claude Haiku 3	Multiple regions	Quick and accurate targeted performance
Claude Haiku 4.5	Requires processing region	Fastest model with near-frontier intelligence
Claude Sonnet 3.5 v2	Limited regions	Updated version with improved capabilities
Claude Sonnet 3.7	Requires processing region	High-performance model with improved reasoning
Claude Sonnet 3	Limited regions	Balance of intelligence and speed
Claude Sonnet 4.5	Requires processing region	Smartest model for complex AI agents and coding
Claude Sonnet 4.6	Requires processing region	Reasoning model with larger context window
Claude Sonnet 4	Requires processing region	High intelligence and balanced performance
GPT 5.4	Limited regions	Latest flagship model with cutting-edge reasoning capabilities; extended thinking always enabled
GPT 5.5	Limited regions	Most capable model with advanced reasoning and multimodal understanding; extended thinking always enabled
Titan Text Embeddings V2	Multiple regions	Text embedding for semantic search

Bring-your-own AI models

You can configure which AI provider powers Appian's generative AI features. By default, Appian uses its own infrastructure. However, you can bring your own provider to gain greater control over security, compliance, and cloud resource management. Appian does not restrict, validate, or select the model when you have external providers configured. The model is determined entirely by the provider you configure, so Appian works with any current or future model that provider offers.

Model	Provider
Claude Opus 4.5	Anthropic
Claude Opus 4.6	Anthropic
Claude Sonnet 4.6	Anthropic
Claude Sonnet 4.5	Anthropic
Claude Haiku 4.5	Anthropic
GPT-5	OpenAI
GPT-5 pro	OpenAI
GPT-5 Codex	OpenAI
GPT-5 mini	OpenAI
GPT-5 nano	OpenAI
GPT-5.1	OpenAI
GPT-5.1 Codex	OpenAI
GPT-5.1 Codex Max	OpenAI
GPT-5.2	OpenAI
GPT-5.2 Codex	OpenAI
GPT-5.3 Codex	OpenAI
GPT-5.4	OpenAI
GPT-5.4 mini	OpenAI
GPT-5.4 nano	OpenAI
GPT-5.5	OpenAI
GPT-4.1	OpenAI
GPT-4.1 mini	OpenAI
GPT-4.1 nano	OpenAI
GPT-4o	OpenAI
GPT-4o mini	OpenAI
Gemini 3.1 Flash-Lite	Google
Gemini 3.5 Flash-Lite	Google
Gemini 3.5 Flash	Google
Gemini 3.7 Flash	Google

PDF processing limits by model

The PDF limits for "Visual elements & text" document processing mode vary by AI model:

Model	Max pages	Max file size
Claude Sonnet 5	100	24 MB
Claude Sonnet 4.6	100	24 MB
Claude Sonnet 4	100	24 MB
Claude Sonnet 4.5	100	24 MB
Claude Haiku 4.5	100	24 MB
Amazon Nova 2 Lite	100	18 MB
All Azure OpenAI GPT models	100	24 MB
All Appian GPT models	100	24 MB
Claude Sonnet 3.5 (legacy)	20	25 MB
Claude Sonnet 3.7 (legacy)	20	25 MB

Administrators can also disable all models from a provider at the environment level to prevent Appian from routing requests to that provider.

Region-restricted AI models

Some AI models are not supported in all regions. Feature availability depends on whether the AI model's selected processing region includes a region where the model is supported.

To ensure AI features continue to work as expected, make sure your environment uses an AI model that supports the processing region for the region where your data is processed. Review your processing region configuration in the Administration Console under AI Services.

AI model deprecation schedule

AI models are periodically deprecated by their respective providers, such as Anthropic and AWS. When this happens, Appian updates the available AI models to ensure the best performance, security, and compliance. Older models may be deprecated as part of this lifecycle. Appian communicates these changes in advance to help you plan transitions.

The following table lists all AI model deprecations, both past and scheduled. The Status column shows whether each deprecation has already taken effect or is still upcoming. Click any column heading to sort the table.

Model	Status	Deprecation date	Replacement	Impacted regions
Claude Haiku 3.5	Deprecated	February 19, 2026		All regions where this model is offered
Claude Sonnet 3.5 v2	Deprecated	March 1, 2026	Latest available Sonnet model	us-east-1 (N. Virginia), us-west-2 (Oregon), eu-central-1 (Frankfurt), eu-central-2 (Zurich)
Claude Sonnet 3.5 v2	Deprecated	June 1, 2026	Latest available Sonnet model	eu-central-2 (Zurich)
Claude Sonnet 3.5 v2	Deprecated	July 30, 2026	Latest available Sonnet model	us-gov-west-1 (GovCloud), us-gov-east-1 (GovCloud), ap-northeast-1 (Tokyo), ap-southeast-2 (Sydney)
Claude Sonnet 3	Deprecated	July 30, 2026	Latest available Sonnet model	ap-northeast-1 (Tokyo), ap-northeast-2 (Seoul), ap-southeast-2 (Sydney), eu-west-1 (Ireland), eu-west-3 (Paris), us-east-1 (N. Virginia), us-west-2 (Oregon)
Claude Haiku 3	Deprecated	September 10, 2026		ap-northeast-1 (Tokyo), ap-southeast-2 (Sydney), eu-central-1 (Frankfurt), eu-west-1 (Ireland), eu-west-2 (London), eu-west-3 (Paris), us-east-1 (N. Virginia), us-west-2 (Oregon), us-gov-west-1 (GovCloud), us-gov-east-1 (GovCloud)
Claude Sonnet 4	Deprecated	October 14, 2026	Latest available Sonnet model	All regions where this model is offered
Claude Haiku 3	Scheduled	November 13, 2026		ap-northeast-2 (Seoul), ap-south-1 (Mumbai), ap-southeast-1 (Singapore), eu-central-2 (Zurich), sa-east-1 (São Paulo), ca-central-1 (Canada)
Claude Sonnet 3.5 v2	Scheduled	November 13, 2026		ap-northeast-2 (Seoul), ap-south-1 (Mumbai), ap-southeast-1 (Singapore)
Claude Sonnet 3.7	Scheduled	November 13, 2026		ap-south-1 (Mumbai), eu-west-2 (London)
Claude Sonnet 3	Scheduled	November 13, 2026		ap-south-1 (Mumbai), ap-southeast-1 (Singapore), ca-central-1 (Canada), eu-west-2 (London), sa-east-1 (São Paulo)
Claude Sonnet 4.5	Scheduled	April 8, 2027		All regions where this model is offered

What happens when an AI model is deprecated

When an AI model is deprecated:

• If the AI skill uses Auto mode: AI skills configured with Auto model selection automatically select the Appian recommended model within the same provider when a model is deprecated. No manual action is required.
• If the AI skill uses a specific model: Existing AI skills using the deprecated model continue to function until the model is fully removed. Appian provides advance notice so you can update your skills to use a replacement model.
• Required actions: Administrators must update all AI skills and agents that use a specific deprecated model before the deprecation deadline to avoid service disruptions. AI skills using Auto mode do not require updates for model deprecations.
• Regional implications: If a replacement model has different regional availability than the deprecated model, you must adjust your processing region configuration.

Prepare for AI model deprecations

To ensure smooth operation of AI features and prepare for AI model changes:

• Enabled AI models: Verify enabled AI models in the Administration Console under AI Services.
• Processing regions: Review processing regions to ensure they include regions that support your required AI models.
• Feature availability: Check feature availability in the Regional Availability page.
• Deprecation timelines: Monitor deprecation timelines in the AI model deprecation schedule section and plan updates accordingly.
• Testing: Test AI skills and agents after making any model or processing region changes.
• Release notes: Review release notes to stay informed about AI model updates and deprecations.

For more information about AI model configuration and availability, see the following pages:

• For configuration, see AI Services.
• For regional availability and processing regions, see Regional Availability.
• For data processing and encryption details, see Security & Compliance.

Feedback