Preview features are deemed stable; however, they remain under active development and may be subject to shorter deprecation windows. Preview features are suitable for early adopters in production environments.

We release preview features so partners and developers can study, test, and integrate them prior to General Availability (GA). For more information, see Forge release phases: EAP, Preview, and GA.

Forge LLMs supports Claude models across three tiers: Haiku, Sonnet, and Opus. You choose the model per request, allowing you to balance latency, capability, and cost for each use case.

Supported models

You can use the list method from the @forge/llm SDK to dynamically fetch the list of supported models and their current status.

Model ID	Tier	Family	Status
`claude-haiku-4-5-20251001`	Haiku	Claude	`ACTIVE`
`claude-sonnet-4-5-20250929`	Sonnet	Claude	`ACTIVE`
`claude-sonnet-4-6`	Sonnet	Claude	`ACTIVE`
`claude-sonnet-5`	Sonnet	Claude	`ACTIVE`
`claude-opus-4-6`	Opus	Claude	`ACTIVE`
`claude-opus-4-7`	Opus	Claude	`ACTIVE`

AI models evolve quickly. Check the status and EOL fields regularly, and update your app before a model reaches end-of-life to avoid disruption. Initially only text input/output is supported; multimodal support may be considered later.

Some models do not support the temperature and top_p sampling parameters. Omit both parameters from requests to any affected model. For the current list of affected models, see the validation rules.

Claude - Opus

Most capable (best for complex, deep reasoning tasks)
Slowest (higher latency due to depth)
Highest cost

Claude - Sonnet

Balanced capability
Moderate speed
Moderate cost

Claude - Haiku

Fast and efficient (best for lightweight or high‑volume tasks)
Lowest cost