DeepInfra Provider Setup in OpenClaw

DeepInfra serves popular open source and frontier models through one OpenAI-compatible endpoint, secured by a single API key. Most OpenAI SDKs can talk to it after you change the base URL.

Install plugin

openclaw plugins install @openclaw/deepinfra-provider
openclaw gateway restart

Get an API key

Sign in at deepinfra.com
Open Dashboard / Keys and create a key, or reuse the one generated automatically

CLI setup

openclaw onboard --deepinfra-api-key <key>

Alternatively, set the environment variable:

export DEEPINFRA_API_KEY="<your-deepinfra-api-key>" # pragma: allowlist secret

Config snippet

{
  env: { vars: { DEEPINFRA_API_KEY: "<your-deepinfra-api-key>" } }, // pragma: allowlist secret
  agents: {
    defaults: {
      model: { primary: "deepinfra/deepseek-ai/DeepSeek-V4-Flash" },
    },
  },
}

Supported surfaces

Chat, image generation, and video generation pull their model lists live from https://api.deepinfra.com/v1/openai/models?sort_by=openclaw&filter=with_meta once DEEPINFRA_API_KEY is set up. This live discovery makes more models available for selection; the default model on each surface remains the static value shown below. Every other surface relies on static catalogs until it adopts the live catalog.

Surface	Default model	OpenClaw config/tool
Chat / model provider	`deepseek-ai/DeepSeek-V4-Flash` (live catalog adds more chat models)	`agents.defaults.model`
Image generation/editing	`black-forest-labs/FLUX-1-schnell` (live catalog adds more `image-gen` models)	`image_generate`, `agents.defaults.mediaModels.image`
Media understanding	`moonshotai/Kimi-K2.5` for images	inbound image understanding
Speech-to-text	`openai/whisper-large-v3-turbo`	inbound audio transcription
Text-to-speech	`hexgrad/Kokoro-82M`	`tts.provider: "deepinfra"`
Video generation	`Pixverse/Pixverse-T2V` (live catalog adds more `video-gen` models)	`video_generate`, `agents.defaults.mediaModels.video`
Memory embeddings	`BAAI/bge-m3`	`memory.search.provider: "deepinfra"`

DeepInfra additionally offers reranking, classification, object-detection, and other native model types. OpenClaw lacks a provider contract for those categories at this time, so the plugin does not register them.

Available models

Once a key is in place, OpenClaw finds DeepInfra models dynamically. Run /models deepinfra or openclaw models list --provider deepinfra to view the current list.

Every model hosted on deepinfra.com is reachable with the deepinfra/ prefix:

deepinfra/deepseek-ai/DeepSeek-V4-Flash
deepinfra/deepseek-ai/DeepSeek-V4-Pro
deepinfra/zai-org/GLM-5.2
deepinfra/stepfun-ai/Step-3.7-Flash
deepinfra/moonshotai/Kimi-K2.7-Code
deepinfra/moonshotai/Kimi-K2.6
deepinfra/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B
deepinfra/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B
...and many more

Notes

Model refs take the form deepinfra/<provider>/<model> (for instance deepinfra/Qwen/Qwen3-Max).
Default chat model: deepinfra/deepseek-ai/DeepSeek-V4-Flash
Base URL: https://api.deepinfra.com/v1/openai
Video generation relies on the OpenAI-compatible async endpoint https://api.deepinfra.com/v1/openai/videos (submit, then poll). A configured baseUrl is respected. On api.deepinfra.com, openclaw doctor --fix converts legacy nativeBaseUrl or /v1/inference values to baseUrl automatically; custom native endpoints are removed with a doctor notice, requiring a manually configured OpenAI-compatible baseUrl. Video generation stops with a clear error (before any request is sent) when baseUrl still points at the retired /v1/inference surface.