Skip to main content
Agent skill

Omni Inference

The core OpenAI-compatible inference endpoints: chat completions, embeddings, images, audio (TTS/STT), moderations, rerank, and the Responses API. The primary integration surface for AI agents.

Get started

Get the complete skill folder, including its scripts and reference files.

When to use this skill

Use Omni Inference when an AI agent needs a reusable SKILL.md workflow for this job: The core OpenAI-compatible inference endpoints: chat completions, embeddings, images, audio (TTS/STT), moderations, rerank, and the Responses API. The primary integration surface for AI agents.

When not to use it

Skip Omni Inference when the task is outside the coding category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.

How to install

Where will you use it?

Install scope
Claude Code folder setup
mkdir -p ~/.claude/skills/omni-inference
# Save SKILL.md to ~/.claude/skills/omni-inference/SKILL.md
  1. Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/omni-inference/SKILL.md. For a shared project, commit the folder at .claude/skills/omni-inference/ instead.
  2. Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
  3. For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.

Full install guide for Claude, Cursor, and Codex →

Try a starter prompt

After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.

Starter prompt
Use the Omni Inference skill for this task: Use Omni Inference when this documented workflow matches the task. Confirm any required inputs or files before making changes.

What supports this recommendation?

71/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.

Documentation
77/100
Practical detail
68/100
Source evidence
63/100
Source signals
72/100
Model compatibility evidence
  • claudeinferred

    The skill text mentions Claude or a closely associated term. This is keyword-derived compatibility; no hands-on model test is recorded.

  • chatgptinferred

    The skill text mentions ChatGPT or a closely associated term. This is keyword-derived compatibility; no hands-on model test is recorded.

  • geminiuntested

    No model-specific signal or recorded compatibility test was found.

  • copilotuntested

    No model-specific signal or recorded compatibility test was found.

  • llamauntested

    No model-specific signal or recorded compatibility test was found.

  • perplexityuntested

    No model-specific signal or recorded compatibility test was found.

  • mistraluntested

    No model-specific signal or recorded compatibility test was found.

  • grokuntested

    No model-specific signal or recorded compatibility test was found.

Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.

What this skill does

<!-- generated by src/lib/agentSkills/generator.ts; manual edits will be overwritten -->

## Overview

The core OpenAI-compatible inference endpoints: chat completions, embeddings, images, audio (TTS/STT), moderations, rerank, and the Responses API. The primary integration surface for AI agents.

## Authentication

All requests require a valid Bearer token or session cookie. Obtain a token via `POST /api/auth/login` or configure `REQUIRE_API_KEY=false` for local development.

## Endpoints

- [`POST /api/v1/session-leases`](references/endpoints.md#post-apiv1session-leases) - [`GET /api/v1/search`](references/endpoints.md#get-apiv1search) - [`POST /api/v1/search`](references/endpoints.md#post-apiv1search) - [`POST /api/v1/chat/completions`](references/endpoints.md#post-apiv1chatcompletions) - [`GET /api/v1/ws`](references/endpoints.md#get-apiv1ws) - [`POST /api/v1/providers/{provider}/chat/completions`](references/endpoints.md#post-apiv1providersproviderchatcompletions) - [`POST /api/v1/api/chat`](references/endpoints.md#post-apiv1apichat) - [`POST /api/v1/messages`](references/endpoints.md#post-apiv1messages) - [`POST /api/v1/messages/count_tokens`](references/endpoints.md#post-apiv1messagescounttokens) - [`POST /api/v1/responses`](references/endpoints.md#post-apiv1responses) - [`POST /api/v1/embeddings`](references/endpoints.md#post-apiv1embeddings) - [`GET /api/v1/multimodal-embeddings`](references/endpoints.md#get-apiv1multimodal-embeddings) - [`POST /api/v1/multimodal-embeddings`](references/endpoints.md#post-apiv1multimodal-embeddings) - [`POST /api/v1/providers/{provider}/embeddings`](references/endpoints.md#post-apiv1providersproviderembeddings) - [`POST /api/v1/images/generations`](references/endpoints.md#post-apiv1imagesgenerations) - [`POST /api/v1/providers/{provider}/images/generations`](references/endpoints.md#post-apiv1providersproviderimagesgenerations) - [`POST /api/v1/audio/speech`](references/endpoints.md#post-apiv1audiospeech) - [`POST /api/v1/audio/transcriptions`](references/endpoints.md#post-apiv1audiotranscriptions) - [`POST /api/v1/moderations`](references/endpoints.md#post-apiv1moderations) - [`POST /api/v1/rerank`](references/endpoints.md#post-apiv1rerank) - [`GET /api/v1`](references/endpoints.md#get-apiv1) - [`GET /api/v1/providers/{provider}/models`](references/endpoints.md#get-apiv1providersprovidermodels) - [`GET /api/v1/management/proxy-subscriptions`](references/endpoints.md#get-apiv1managementproxy-subscriptions) - [`POST /api/v1/management/proxy-subscriptions`](references/endpoints.md#post-apiv1managementproxy-subscriptions) - [`GET /api/v1/management/proxy-subscriptions/{id}`](references/endpoints.md#get-apiv1managementproxy-subscriptionsid) - [`PATCH /api/v1/management/proxy-subscriptions/{id}`](references/endpoints.md#patch-apiv1managementproxy-subscriptionsid) - [`DELETE /api/v1/management/proxy-subscriptions/{id}`](references/endpoints.md#delete-apiv1managementproxy-subscriptionsid) - [`GET /api/v1/management/proxy-subscriptions/{id}/nodes`](references/endpoints.md#get-apiv1managementproxy-subscriptionsidnodes) - [`POST /api/v1/management/proxy-subscriptions/{id}/refresh`](references/endpoints.md#post-apiv1managementproxy-subscriptionsidrefresh) - [`POST /api/v1/ocr`](references/endpoints.md#post-apiv1ocr) - [`POST /api/v1/audio/translations`](references/endpoints.md#post-apiv1audiotranslations) - [`GET /api/v1/voices`](references/endpoints.md#get-apiv1voices) - [`POST /api/v1/speech-to-text`](references/endpoints.md#post-apiv1speech-to-text) - [`POST /api/v1/text-to-speech/{voiceId}`](references/endpoints.md#post-apiv1text-to-speechvoiceid) - [`POST /api/v1/sound-generation`](references/endpoints.md#post-apiv1sound-generation) - [`POST /api/v1/music`](references/endpoints.md#post-apiv1music) - [`POST /api/v1/music/stream`](references/endpoints.md#post-apiv1musicstream) - [`POST /api/v1/music/detailed`](references/endpoints.md#post-apiv1musicdetailed) - [`POST /api/v1/music/detailed/stream`](references/endpoints.md#post-apiv1musicdetailedstream) - [`POST /api/v1/music/plan`](references/endpoints.md#post-apiv1musicplan) - [`GET /api/v1/explain/routing`](references/endpoints.md#get-apiv1explainrouting) - [`GET /api/v1/providers/suggested-models`](references/endpoints.md#get-apiv1providerssuggested-models) - [`GET /api/v1/provider-plugin-manifest`](references/endpoints.md#get-apiv1provider-plugin-manifest) - [`GET /api/v1/{omnirouteCatchAll}`](references/endpoints.md#get-apiv1omniroutecatchall) - [`POST /api/v1/{omnirouteCatchAll}`](references/endpoints.md#post-apiv1omniroutecatchall) - [`PUT /api/v1/{omnirouteCatchAll}`](references/endpoints.md#put-apiv1omniroutecatchall) - [`PATCH /api/v1/{omnirouteCatchAll}`](references/endpoints.md#patch-apiv1omniroutecatchall) - [`DELETE /api/v1/{omnirouteCatchAll}`](references/endpoints.md#delete-apiv1omniroutecatchall) - [`GET /api/v1/accounts/{id}/limits`](references/endpoints.md#get-apiv1accountsidlimits) - [`PUT /api/v1/accounts/{id}/limits`](references/endpoints.md#put-apiv1accountsidlimits) - [`GET /api/v1/agents/credentials`](references/endpoints.md#get-apiv1agentscredentials) - [`POST /api/v1/agents/credentials`](references/endpoints.md#post-apiv1agentscredentials) - [`GET /api/v1/agents/health`](references/endpoints.md#get-apiv1agentshealth) - [`GET /api/v1/agents/tasks`](references/endpoints.md#get-apiv1agentstasks) - [`POST /api/v1/agents/tasks`](references/endpoints.md#post-apiv1agentstasks) - [`DELETE /api/v1/agents/tasks`](references/endpoints.md#delete-apiv1agentstasks) - [`GET /api/v1/agents/tasks/{id}`](references/endpoints.md#get-apiv1agentstasksid) - [`POST /api/v1/agents/tasks/{id}`](references/endpoints.md#post-apiv1agentstasksid) - [`DELETE /api/v1/agents/tasks/{id}`](references/endpoints.md#delete-apiv1agentstasksid) - [`POST /api/v1/antigravity`](references/endpoints.md#post-apiv1antigravity) - [`GET /api/v1/auto-combo/{channel}/candidates`](references/endpoints.md#get-apiv1auto-combochannelcandidates) - [`GET /api/v1/batches`](references/endpoints.md#get-apiv1batches) - [`POST /api/v1/batches`](references/endpoints.md#post-apiv1batches) - [`GET /api/v1/batches/{id}`](references/endpoints.md#get-apiv1batchesid) - [`DELETE /api/v1/batches/{id}`](references/endpoints.md#delete-apiv1batchesid) - [`POST /api/v1/batches/{id}/cancel`](references/endpoints.md#post-apiv1batchesidcancel) - [`DELETE /api/v1/batches/delete-completed`](references/endpoints.md#delete-apiv1batchesdelete-completed) - [`POST /api/v1/classify`](references/endpoints.md#post-apiv1classify) - [`GET /api/v1/combos`](references/endpoints.md#get-apiv1combos) - [`POST /api/v1/completions`](references/endpoints.md#post-apiv1completions) - [`GET /api/v1/files`](references/endpoints.md#get-apiv1files) - [`POST /api/v1/files`](references/endpoints.md#post-apiv1files) - [`GET /api/v1/files/{id}`](references/endpoints.md#get-apiv1filesid) - [`DELETE /api/v1/files/{id}`](references/endpoints.md#delete-apiv1filesid) - [`GET /api/v1/files/{id}/content`](references/endpoints.md#get-apiv1filesidcontent) - [`POST /api/v1/images/edits`](references/endpoints.md#post-apiv1imagesedits) - [`GET /api/v1/images/upscale`](references/endpoints.md#get-apiv1imagesupscale) - [`POST /api/v1/images/upscale`](references/endpoints.md#post-apiv1imagesupscale) - [`GET /api/v1/media/{id}`](references/endpoints.md#get-apiv1mediaid) - [`POST /api/v1/issues/report`](references/endpoints.md#post-apiv1issuesreport) - [`GET /api/v1/management/proxies`](references/endpoints.md#get-apiv1managementproxies) - [`POST /api/v1/management/proxies`](references/endpoints.md#post-apiv1managementproxies) - [`PATCH /api/v1/management/proxies`](references/endpoints.md#patch-apiv1managementproxies) - [`DELETE /api/v1/management/proxies`](references/endpoints.md#delete-apiv1managementproxies) - [`GET /api/v1/management/proxies/assignments`](references/endpoints.md#get-apiv1managementproxiesassignments) - [`PUT /api/v1/management/proxies/assignments`](references/endpoints.md#put-apiv1managementproxiesassignments) - [`PUT /api/v1/management/proxies/bulk-assign`](references/endpoints.md#put-apiv1managementproxiesbulk-assign) - [`GET /api/v1/management/proxies/health`](references/endpoints.md#get-apiv1managementproxieshealth) - [`GET /api/v1/me/status`](references/endpoints.md#get-apiv1mestatus) - [`GET /api/v1/me/sessions`](references/endpoints.md#get-apiv1mesessions) - [`GET /api/v1/me/sessions/{id}`](references/endpoints.md#get-apiv1mesessionsid) - [`GET /api/v1/muse-code/models`](references/endpoints.md#get-apiv1muse-codemodels) - [`GET /api/v1/music/generations`](references/endpoints.md#get-apiv1musicgenerations) - [`POST /api/v1/music/generations`](references/endpoints.md#post-apiv1musicgenerations) - [`GET /api/v1/providers/{provider}/limits`](references/endpoints.md#get-apiv1providersproviderlimits) - [`PUT /api/v1/providers/{provider}/limits`](references/endpoints.md#put-apiv1providersproviderlimits) - [`GET /api/v1/quota-recoveries/{id}`](references/endpoints.md#get-apiv1quota-recoveriesid) - [`DELETE /api/v1/quota-recoveries/{id}`](references/endpoints.md#delete-apiv1quota-recoveriesid) - [`GET /api/v1/quotas/check`](references/endpoints.md#get-apiv1quotascheck) - [`GET /api/v1/registered-keys`](references/endpoints.md#get-apiv1registered-keys) - [`POST /api/v1/registered-keys`](references/endpoints.md#post-apiv1registered-keys) - [`GET /api/v1/registered-keys/{id}`](references/endpoints.md#get-apiv1registered-keysid) - [`DELETE /api/v1/registered-keys/{id}`](references/endpoints.md#delete-apiv1registered-keysid) - [`POST /api/v1/registered-keys/{id}/revoke`](references/endpoints.md#post-apiv1registered-keysidrevoke) - [`POST /api/v1/relay/chat/completions`](references/endpoints.md#post-apiv1relaychatcompletions) - [`POST /api/v1/relay/chat/completions/bifrost`](references/endpoints.md#post-apiv1relaychatcompletionsbifrost) - [`POST /api/v1/responses/{path}`](references/endpoints.md#post-apiv1responsespath) - [`GET /api/v1/search/analytics`](references/endpoints.md#get-apiv1searchanalytics) - [`POST /api/v1/segment`](references/endpoints.md#post-apiv1segment) - [`GET /api/v1/systemone`](references/endpoints.md#get-apiv1systemone) - [`POST /api/v1/systemone`](references/endpoints.md#post-apiv1systemone) - [`GET /api/v1/video-bridge/drilldown`](references/endpoints.md#get-apiv1video-bridgedrilldown) - [`DELETE /api/v1/video-bridge/drilldown`](references/endpoints.md#delete-apiv1video-bridgedrilldown) - [`GET /api/v1/videos/generations`](references/endpoints.md#get-apiv1videosgenerations) - [`POST /api/v1/videos/generations`](references/endpoints.md#post-apiv1videosgenerations) - [`GET /api/v1/vscode/{token}`](references/endpoints.md#get-apiv1vscodetoken) - [`POST /api/v1/vscode/{token}/api/chat`](references/endpoints.md#post-apiv1vscodetokenapichat) - [`POST /api/v1/vscode/{token}/api/show`](references/endpoints.md#post-apiv1vscodetokenapishow) - [`GET /api/v1/vscode/{token}/api/tags`](references/endpoints.md#get-apiv1vscodetokenapitags) - [`GET /api/v1/vscode/{token}/api/version`](references/endpoints.md#get-apiv1vscodetokenapiversion) - [`POST /api/v1/vscode/{token}/chat/completions`](references/endpoints.md#post-apiv1vscodetokenchatcompletions) - [`GET /api/v1/vscode/{token}/combos`](references/endpoints.md#get-apiv1vscodetokencombos) - [`GET /api/v1/vscode/{token}/models`](references/endpoints.md#get-apiv1vscodetokenmodels) - [`POST /api/v1/vscode/{token}/responses`](references/endpoints.md#post-apiv1vscodetokenresponses) - [`POST /api/v1/vscode/{token}/v1/chat/completions`](references/endpoints.md#post-apiv1vscodetokenv1chatcompletions) - [`GET /api/v1/vscode/{token}/v1/models`](references/endpoints.md#get-apiv1vscodetokenv1models) - [`GET /api/v1/vscode/combos/{token}/{{slug}}`](references/endpoints.md#get-apiv1vscodecombostokenslug) - [`POST /api/v1/vscode/combos/{token}/{{slug}}`](references/endpoints.md#post-apiv1vscodecombostokenslug) - [`GET /api/v1/vscode/raw/{token}`](references/endpoints.md#get-apiv1vscoderawtoken) - [`POST /api/v1/vscode/raw/{token}/api/chat`](references/endpoints.md#post-apiv1vscoderawtokenapichat) - [`POST /api/v1/vscode/raw/{token}/api/show`](references/endpoints.md#post-apiv1vscoderawtokenapishow) - [`GET /api/v1/vscode/raw/{token}/api/tags`](references/endpoints.md#get-apiv1vscoderawtokenapitags) - [`GET /api/v1/vscode/raw/{token}/api/version`](references/endpoints.md#get-apiv1vscoderawtokenapiversion) - [`POST /api/v1/vscode/raw/{token}/chat/completions`](references/endpoints.md#post-apiv1vscoderawtokenchatcompletions) - [`GET /api/v1/vscode/raw/{token}/combos`](references/endpoints.md#get-apiv1vscoderawtokencombos) - [`GET /api/v1/vscode/raw/{token}/models`](references/endpoints.md#get-apiv1vscoderawtokenmodels) - [`POST /api/v1/vscode/raw/{token}/responses`](references/endpoints.md#post-apiv1vscoderawtokenresponses) - [`POST /api/v1/vscode/raw/{token}/v1/chat/completions`](references/endpoints.md#post-apiv1vscoderawtokenv1chatcompletions) - [`GET /api/v1/vscode/raw/{token}/v1/models`](references/endpoints.md#get-apiv1vscoderawtokenv1models) - [`POST /api/v1/web/fetch`](references/endpoints.md#post-apiv1webfetch) ## Payloads

See the full OpenAPI specification at `GET /api/openapi/spec` or `docs/openapi.yaml` for detailed request/response schemas.

<!-- skill:custom-start --> <!-- Aggregated from: omniroute-chat, omniroute-image, omniroute-tts, omniroute-stt, omniroute-embeddings, omniroute-web-search, omniroute-web-fetch -->

## Chat completions

Requires `OMNIROUTE_URL` and `OMNIROUTE_KEY`. See [entry-point SKILL](https://raw.githubusercontent.com/cheaper-inference/OmniRoute/main/skills/omni-auth/SKILL.md) for setup.

### Endpoints

- `POST $OMNIROUTE_URL/v1/chat/completions` — OpenAI format - `POST $OMNIROUTE_URL/v1/messages` — Anthropic Messages format - `POST $OMNIROUTE_URL/v1/responses` — OpenAI Responses API

### Discover

```bash curl $OMNIROUTE_URL/v1/models | jq '.data[].id' ```

Combos (e.g. `auto`, `cost-optimized`, `subscription`) auto-fallback through multiple providers.

### OpenAI format example

```bash curl -X POST $OMNIROUTE_URL/v1/chat/completions \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "claude-opus-4-7", "messages": [{"role": "user", "content": "Refactor this function"}], "stream": true }' ```

### Anthropic format example

```bash curl -X POST $OMNIROUTE_URL/v1/messages \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "anthropic-version: 2023-06-01" \ -H "Content-Type: application/json" \ -d '{ "model": "claude-opus-4-7", "max_tokens": 4096, "messages": [{"role": "user", "content": "Hi"}] }' ```

### Tool use

Supports OpenAI `tools` array and Anthropic `tools` block. Tool results auto-compressed via RTK (47 filters: git-diff, grep, test-jest, terraform-plan, docker-logs, etc.) — 20-40% token savings. Disable per-request with `x-omniroute-compression: off` header.

### Reasoning / thinking

Anthropic extended thinking and OpenAI Responses reasoning blocks are forwarded verbatim. Cached automatically via reasoning cache.

### Errors

- `401` → invalid API key - `400 Invalid model format` → model string not parseable; check `/v1/models` - `503 provider_circuit_open` → provider circuit breaker tripped; honor `Retry-After` or use combo - `429 rate_limited` → honor `Retry-After`; consider using a combo for auto-fallback

## Image generation

Requires `OMNIROUTE_URL` and `OMNIROUTE_KEY`. See [entry-point SKILL](https://raw.githubusercontent.com/cheaper-inference/OmniRoute/main/skills/omni-auth/SKILL.md) for setup.

### Endpoints

- `POST $OMNIROUTE_URL/v1/images/generations` — Text-to-image - `POST $OMNIROUTE_URL/v1/images/edits` — Image edit (mask) - `POST $OMNIROUTE_URL/v1/images/upscale` — Upscale

### Discover

```bash curl "$OMNIROUTE_URL/v1/models?kind=image" | jq '.data[]' ```

### Generate example

```bash curl -X POST $OMNIROUTE_URL/v1/images/generations \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "dall-e-3", "prompt": "a red bicycle on a wet street, photoreal", "n": 1, "size": "1024x1024", "response_format": "b64_json" }' ```

Response: `{ created, data: [{ url? or b64_json, revised_prompt }] }`

### Errors

- `400 content_policy_violation` → blocked by provider safety - `503` → provider unavailable; try another model from `/v1/models?kind=image`

## Text-to-speech

Requires `OMNIROUTE_URL` and `OMNIROUTE_KEY`. See [entry-point SKILL](https://raw.githubusercontent.com/cheaper-inference/OmniRoute/main/skills/omni-auth/SKILL.md) for setup.

### Endpoint

- `POST $OMNIROUTE_URL/v1/audio/speech` — returns binary audio (mp3/opus/wav/flac)

### Discover

```bash curl "$OMNIROUTE_URL/v1/models?kind=audio" | jq '.data[]' curl "$OMNIROUTE_URL/v1/voices" | jq ```

### Example

```bash curl -X POST $OMNIROUTE_URL/v1/audio/speech \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "tts-1", "input": "Hello from OmniRoute.", "voice": "alloy", "response_format": "mp3" }' --output speech.mp3 ```

### Voices

Voice names vary by provider. List them with `GET /v1/voices`. Common OpenAI voices: `alloy`, `echo`, `fable`, `onyx`, `nova`, `shimmer`.

### Errors

- `503` → provider unavailable; try another model from `/v1/models?kind=audio`

## Speech-to-text

Requires `OMNIROUTE_URL` and `OMNIROUTE_KEY`. See [entry-point SKILL](https://raw.githubusercontent.com/cheaper-inference/OmniRoute/main/skills/omni-auth/SKILL.md) for setup.

### Endpoints

- `POST $OMNIROUTE_URL/v1/audio/transcriptions` — multipart upload, returns text - `POST $OMNIROUTE_URL/v1/audio/translations` — transcribe + translate to English

### Discover

```bash curl "$OMNIROUTE_URL/v1/models?kind=audio" | jq '.data[]' ```

### Example

```bash curl -X POST $OMNIROUTE_URL/v1/audio/transcriptions \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -F "[email protected]" \ -F "model=whisper-1" \ -F "response_format=verbose_json" ```

Response: `{ text, language, duration, segments?:[{ start, end, text }] }`

### Supported formats

Audio: `mp3`, `mp4`, `mpeg`, `mpga`, `m4a`, `wav`, `webm`. Response formats: `json`, `text`, `srt`, `verbose_json`, `vtt`.

### Errors

- `400 file_too_large` → exceeds provider limit (usually 25MB) - `503` → provider unavailable; try another model from `/v1/models?kind=audio`

## Embeddings

Requires `OMNIROUTE_URL` and `OMNIROUTE_KEY`. See [entry-point SKILL](https://raw.githubusercontent.com/cheaper-inference/OmniRoute/main/skills/omni-auth/SKILL.md) for setup.

### Endpoint

- `POST $OMNIROUTE_URL/v1/embeddings`

### Discover

```bash curl "$OMNIROUTE_URL/v1/models?kind=embedding" | jq '.data[]' ```

### Example

```bash curl -X POST $OMNIROUTE_URL/v1/embeddings \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "Content-Type: application/json" \ -d '{ "model": "text-embedding-3-large", "input": ["first text", "second text"], "encoding_format": "float" }' ```

Response: `{ data:[{ embedding:[...], index }], usage:{ prompt_tokens, total_tokens } }`

### Batch input

`input` accepts a string or array of strings (up to provider batch limit, typically 2048 items).

### Errors

- `503` → provider unavailable; try another model from `/v1/models?kind=embedding`

## Web search

Setup: see `omni-auth`. Search: `POST $OMNIROUTE_URL/v1/search`. List providers: `GET $OMNIROUTE_URL/v1/search`.

```bash curl -X POST "$OMNIROUTE_URL/v1/search" \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "Content-Type: application/json" \ -d '{"query":"OmniRoute release notes","provider":"tavily","max_results":5,"search_type":"web"}' ```

| Field | Notes | | --- | --- | | `query` | Required. 1 to 500 characters. | | `provider` | Alias or id, for example `tavily` or `tavily-search`. If you omit it, OmniRoute walks the providers by cost and uses the first one that has credentials. | | `max_results` | 1 to 100. Default 5. | | `search_type` | `web` (default), `news`, `x`, `finance`. Not every provider supports every type. | | `time_range` | `any`, `hour`, `day`, `week`, `month`, `year`. | | `filters` | `include_domains`, `exclude_domains` (20 each), `safe_search`. | | `provider_options` | Passed to the provider. Tavily: `{"tavily":{"search_depth":"advanced"}}`. Google PSE needs `{"cx":"<engine id>"}`. |

Do not send `model`. OmniRoute ignores it. It does not select a provider.

Response (values are examples):

```json { "provider": "tavily-search", "query": "...", "results": [{ "title": "...", "url": "...", "snippet": "...", "position": 1, "score": 0.9, "citation": { "provider": "tavily-search", "retrieved_at": "...", "rank": 1 } }], "answer": null, "usage": { "queries_used": 1, "search_cost_usd": 0.008 }, "metrics": { "response_time_ms": 850, "upstream_latency_ms": 700, "total_results_available": 12 }, "errors": [] } ```

`answer` is always `null` today. Read `results[].snippet`. Top-level `include_answer` and `search_depth` are ignored. Errors: `400 Unknown search provider: <id>`, `400 ... does not support search_type: <type>`, `400 No credentials configured for search provider: <id>`.

## Web fetch

`POST $OMNIROUTE_URL/v1/web/fetch`

```bash curl -X POST "$OMNIROUTE_URL/v1/web/fetch" \ -H "Authorization: Bearer $OMNIROUTE_KEY" \ -H "Content-Type: application/json" \ -d '{"url":"https://example.com","provider":"jina-reader","format":"markdown","include_metadata":true}' ```

| Field | Notes | | --- | --- | | `url` | Required. A valid `http` or `https` URL. | | `provider` | `firecrawl`, `jina-reader`, `tavily-search`, `tinyfish`, `anysearch-search`, `context7`, `nimble-search`, `parallel-search`. If you omit it, OmniRoute uses the first provider that has credentials, in this order. `context7` is never automatic. An explicit provider has no fallback. | | `format` | `markdown` (default), `html`, `links`, `screenshot`. `text` returns 400. Some providers support only `markdown`. | | `depth`, `wait_for_selector` | Used by Firecrawl. Other providers ignore them. | | `include_metadata` | `true` adds `metadata.title` and `metadata.description`. |

Response (values are examples):

```json { "provider": "jina-reader", "url": "https://example.com", "content": "# Example Domain ...", "links": [], "metadata": { "title": "Example Domain", "description": null }, "screenshot_url": null } ```

`content` is a string. Errors: `400 No credentials configured for web-fetch provider: <id>`, `401 Authentication required` (only when a key is required), `429 All configured web-fetch providers are rate limited or quota-exhausted`, `web_fetch_error` with the upstream status (default 502). The `/v1/models` catalog has no web-fetch entries. The eight ids above are the full set.

<!-- skill:custom-end -->

Intended uses

  • Use Omni Inference when this documented workflow matches the task.

Implementation notes

  • Review the source instructions and adapt inputs before running the workflow.

Try it, then tell us how it went

A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.