Creates, reads, updates, and deletes Arize AI integrations that store LLM provider credentials used by evaluators and other Arize features. Supports any LLM provider (e.g. OpenAI, Anthropic, Azure OpenAI, AWS Bedrock, Vertex AI, Gemini, NVIDIA NIM). Use when the user mentions AI integration, LLM pro

Arize AI Integration Skill SPACE — space flags accept a space name (e.g., my workspace ) or a base64 space ID (e.g., U3BhY2U6... ). Resolve a known name with ax spaces get NAME ; ax spaces list is for browsing when the name is unknown. Note: ai integrations create does not accept space — AI integrations are account scoped. Use space only with list , get , update , and delete . Concepts AI Integration = stored LLM provider credentials registered in Arize; used by evaluators to call a judge model and by other Arize features that need to invoke an LLM on your behalf Provider = the LLM service backing the integration (e.g., OPEN AI , ANTHROPIC , AWS BEDROCK ) Integration ID = a base64 encoded global identifier for an integration (e.g., TGxtSW50ZWdyYXRpb246MTI6YUJjRA== ); required for evaluator creation and other downstream operations Scoping = visibility rules controlling which spaces or users can use an integration Auth type = how Arize authenticates with the provider: DEFAULT (provider API key), PROXY WITH HEADERS (proxy via custom headers), BEARER TOKEN (bearer token auth), or OAUTH2 CLIENT CREDENTIALS Prerequisites Proceed directly with the task — run the ax command you need. Do NOT check versions, env vars, or profiles upfront. If an ax command fails, troubleshoot based on the error: command not found or version error → see [references/ax setup.md](references/ax setup.md) 401 Unauthorized / missing API key → run ax profiles show to inspect the current profile. If the profile is missing or the API key is wrong, follow [references/ax profiles.md](references/ax profiles.md) to create/update it. If the user doesn't have their key, direct them to https://app.arize.com/admin API Keys Space unknown → see [Space](references/ax profiles.md space) LLM provider call fails (missing provider credentials) → run ax ai integrations list space SPACE to check for platform managed credentials. If none exist: Preferred: Give the user the exact ax ai integrations create command from Supported Providers below (reference env var names like $OPENAI API KEY , never raw values). Ask them to export the provider key in their own terminal and run the command there — never paste the key into chat. Do not run the create command yourself unless the var is already exported in this terminal session. Bedrock/Vertex: use provider metadata instead of api key (see Supported Providers below). Security: Never read .env files or search the filesystem for credentials. Exception: the non secret ARIZE SPACE ID line (see [Space](references/ax profiles.md space)). Use ax profiles for Arize credentials and ax ai integrations for LLM provider keys. Never ask the user to paste secrets into chat. For missing credentials, see [references/ax profiles.md](references/ax profiles.md). List AI Integrations List all integrations accessible in a space: Filter by name (case insensitive substring match): Paginate large result sets: Run ax ai integrations list help for flags. When the integration name or ID is known, use ax ai integrations get (below) instead of paging the list; list is for browsing. Response fields: Field Description id Base64 integration ID — copy this for downstream commands name Human readable name provider LLM provider enum (see Supported Providers below) has api key true if credentials are stored model names Allowed model list, or null if all models are enabled enable default models Whether default models for this provider are allowed function calling enabled Whether tool/function calling is enabled auth type Authentication method: DEFAULT , PROXY WITH HEADERS , BEARER TOKEN , or OAUTH2 CLIENT CREDENTIALS Get a Specific Integration Use this to inspect an integration's full configuration or to confirm its ID after creation. Create an AI Integration Before creating, always list integrations first — the user may already have a suitable one: If no suitable integration exists, create one. The required flags depend on the provider. Every create command follows this shape; the provider specific flags come from the table below: Every integration except LiteLLM, Fireworks AI and Together AI needs a model source: pass enable default models (Arize's default model list for the provider) and/or one or more model name flags. Without either, the server rejects the create with 400: An integration must have at least one model available . For a provider's full example and notes (Bedrock/Vertex metadata, base URLs, flags to avoid), open only that provider's file from the Details column below. Supported Providers Provider Required extra flags ("model source" = enable default models or model name ) Key env var Details OPEN AI api key <key , model source $OPENAI API KEY [provider openai.md](references/provider openai.md) ANTHROPIC api key <key , model source $ANTHROPIC API KEY [provider anthropic.md](references/provider anthropic.md) AZURE OPEN AI api key <key , base url <azure endpoint , model name <deployment $AZURE OPENAI API KEY [provider azure openai.md](references/provider azure openai.md) AWS BEDROCK provider metadata '{"role arn": "<arn "}' (no API key), model source — [provider aws bedrock.md](references/provider aws bedrock.md) VERTEX AI provider metadata '{"project id": "<gcp project ", "location": "<region ", "project access label": "<label "}' (no API key), model source — [provider vertex ai.md](references/provider vertex ai.md) GEMINI api key <key , model source $GEMINI API KEY [provider gemini.md](references/provider gemini.md) NVIDIA NIM model source; api key and base url optional (default: NVIDIA hosted) $NVIDIA API KEY [provider nvidia nim.md](references/provider nvidia nim.md) CUSTOM base url <endpoint , model source $CUSTOM LLM API KEY [provider custom.md](references/provider custom.md) LITELLM base url <endpoint , api key <key $LITELLM API KEY [provider litellm.md](references/provider litellm.md) FIREWORKS api key <key only — no base url or headers $FIREWORKS API KEY [provider fireworks.md](references/provider fireworks.md) TOGETHER AI api key <key only — no base url or headers $TOGETHER API KEY [provider together ai.md](references/provider together ai.md) Run ax ai integrations create help for optional flags. Pair headers with auth type PROXY WITH HEADERS when the provider sits behind a proxy. After creation Capture the returned integration ID (e.g., TGxtSW50ZWdyYXRpb246MTI6YUJjRA== ) — it is needed for evaluator creation and other downstream commands. If you missed it, retrieve it: Update an AI Integration update is a partial update — only the flags you provide are changed. Omitted fields stay as is. Add space SPACE when using a name instead of ID. Any flag accepted by create can be passed to update . scopings flag: Controls which spaces can use this integration. Accepts a JSON array of scoping rules. Replaces all existing scopings on update. Use ax spaces list o json to find space global IDs. Delete an AI Integration Warning: Deletion is permanent. Evaluators that reference this integration will no longer be able to run. Omit force to get a confirmation prompt instead of deleting immediately. ax integrations (LLM and agent integrations) ax integrations manages LLM and agent integrations in one group ( list , get , delete , create llm agent , update llm agent ). Run ax integrations <sub help for flags. Choosing a create command: use ax ai integrations create (the examples above) for AZURE OPEN AI or when auth type is needed; ax integrations create llm accepts neither. Use ax integrations create llm when scopings must be set at create time ( scopings ). create llm provider flags differ from ax ai integrations create : Bedrock and Vertex AI take dedicated flags instead of provider metadata . Run ax integrations create llm help for the per provider list. Agent endpoints: use ax integrations create agent to register an agent endpoint; run its help for flags. Provider is immutable in ax integrations update llm ; delete and recreate to change it. type : list returns every type merged; filter with type LLM or type AGENT . get and delete by name need type because names are unique only per type; IDs work without it. Troubleshooting Problem Solution ax: command not found See [references/ax setup.md](references/ax setup.md) 401 Unauthorized API key may not have access to this space. Verify key and space ID at https://app.arize.com/admin API Keys No profile found Run ax profiles show expand ; set ARIZE API KEY env var or write ~/.arize/config.toml Integration not found Check the name or ID with ax ai integrations get NAME OR ID space SPACE ; use ax ai integrations list space SPACE name TEXT only to browse for the right name has api key: false after create Credentials were not saved — re run update with the correct api key or provider metadata Evaluator runs fail with LLM errors Check integration credentials with ax ai integrations get INT ID ; rotate the API key if needed Invalid value for ' provider' with FIREWORKS or TOGETHER AI The installed ax CLI predates these providers — upgrade the ax CLI (see [references/ax setup.md](references/ax setup.md)) An integration must have at least one model available Add enable default models or at least one model name (required for every provider except LiteLLM, Fireworks AI and Together AI) Integration has the wrong provider ax ai integrations update NAME OR ID provider NEW PROVIDER changes it, but other fields are kept as is — pass the new provider's required flags (API key, base URL, model source) in the same call. When the two providers need very different settings, deleting and recreating is cleaner Related Skills arize evaluator : Create LLM as judge evaluators that use an AI integration → use arize evaluator arize experiment : Run experiments that use evaluators backed by an AI integration → use arize experiment Save Credentials for Future Use See [references/ax profiles.md](references/ax profiles.md) § Save Credentials for Future Use.

arize-ai-provider-integration