Task guides
One guide per task the source names: language (chat), classification, image, video, audio, and document processing as the combination of vision and language. Each says whether the task is synchronous (POST /v1/tasks) or asynchronous (POST /v1/jobs), and which route carries it. Embeddings and rerank have routes of their own, POST /v1/embeddings and POST /v1/rerank, described on the Tasks reference, and the image tasks OpenAI's image routes, on Images.
| Task | Route | Mode |
|---|---|---|
| Language (chat) | POST /v1/chat/completions | Synchronous, streamed or not, on POST /v1/chat/completions. |
| Classification | POST /v1/tasks, POST /v1/jobs | Synchronous on POST /v1/tasks; as a job on POST /v1/jobs for a large batch. |
| Image generation and editing | POST /v1/tasks, POST /v1/images/generations, POST /v1/images/edits | Synchronous: the images inline in the answer, billed per image at its size. |
| Video | POST /v1/tasks, POST /v1/jobs | Asynchronous only, on POST /v1/jobs. |
| Audio | POST /v1/tasks, POST /v1/jobs | Synchronous on POST /v1/tasks within the synchronous limits; POST /v1/jobs for long media. |
| Document processing | POST /v1/tasks, POST /v1/jobs | The language step on POST /v1/chat/completions with image_url parts; the specialist steps (classification, detection, embedding) on POST /v1/tasks. |
- Task
- Language (chat)
- Route
POST /v1/chat/completions- Mode
- Synchronous, streamed or not, on
POST /v1/chat/completions.
- Task
- Classification
- Route
POST /v1/tasks,POST /v1/jobs- Mode
- Synchronous on
POST /v1/tasks; as a job onPOST /v1/jobsfor a large batch.
- Route
POST /v1/tasks,POST /v1/images/generations,POST /v1/images/edits- Mode
- Synchronous: the images inline in the answer, billed per image at its size.
- Task
- Video
- Route
POST /v1/tasks,POST /v1/jobs- Mode
- Asynchronous only, on
POST /v1/jobs.
- Task
- Audio
- Route
POST /v1/tasks,POST /v1/jobs- Mode
- Synchronous on
POST /v1/taskswithin the synchronous limits;POST /v1/jobsfor long media.
- Route
POST /v1/tasks,POST /v1/jobs- Mode
- The language step on
POST /v1/chat/completionswithimage_urlparts; the specialist steps (classification, detection, embedding) onPOST /v1/tasks.
| Model type | Shape | Route |
|---|---|---|
| Language model, text in and out | OpenAI, or Anthropic's Messages | POST /v1/chat/completions, POST /v1/responses, POST /v1/messages |
| Language model with image input (vision) | OpenAI image_url parts, Anthropic image blocks | The same three routes, on a model that advertises vision |
| Embeddings | OpenAI | POST /v1/embeddings, or POST /v1/tasks with task embedding |
| Rerank | Cohere | POST /v1/rerank, or POST /v1/tasks with task rerank |
| Classification | HIPS task | POST /v1/tasks |
| Image generation and editing | OpenAI's image routes, or HIPS task | POST /v1/images/generations (JSON) and POST /v1/images/edits (multipart), or POST /v1/tasks with image_generation or image_edit |
| Transcription, speech | OpenAI's routes, not served yet | POST /v1/audio/transcriptions and /v1/audio/speech answer 400 type_not_served until the type is served |
| Detection, translation, video generation | HIPS task, not served yet | The task API's types, served once priced; video generation needs the job routes, not built yet |
| Customer-owned language model | OpenAI | POST /v1/chat/completions with model: dep_... |
| Customer-owned model of any other type | HIPS task | POST /v1/tasks with deployment: dep_... |
- Model type
- Language model, text in and out
- Shape
- OpenAI, or Anthropic's Messages
- Route
POST /v1/chat/completions,POST /v1/responses,POST /v1/messages
- Model type
- Language model with image input (vision)
- Shape
- OpenAI
image_urlparts, Anthropicimageblocks - Route
- The same three routes, on a model that advertises
vision
- Model type
- Embeddings
- Shape
- OpenAI
- Route
POST /v1/embeddings, orPOST /v1/taskswith taskembedding
- Model type
- Rerank
- Shape
- Cohere
- Route
POST /v1/rerank, orPOST /v1/taskswith taskrerank
- Model type
- Classification
- Shape
- HIPS task
- Route
POST /v1/tasks
- Model type
- Image generation and editing
- Shape
- OpenAI's image routes, or HIPS task
- Route
POST /v1/images/generations(JSON) andPOST /v1/images/edits(multipart), orPOST /v1/taskswithimage_generationorimage_edit
- Model type
- Transcription, speech
- Shape
- OpenAI's routes, not served yet
- Route
POST /v1/audio/transcriptionsand/v1/audio/speechanswer 400type_not_serveduntil the type is served
- Model type
- Detection, translation, video generation
- Shape
- HIPS task, not served yet
- Route
- The task API's types, served once priced; video generation needs the job routes, not built yet
- Model type
- Customer-owned language model
- Shape
- OpenAI
- Route
POST /v1/chat/completionswithmodel: dep_...
- Model type
- Customer-owned model of any other type
- Shape
- HIPS task
- Route
POST /v1/taskswithdeployment: dep_...
Note
The job routes are not built yet: POST /v1/jobs and GET /v1/jobs/{id} answer 404 route_not_found, so the asynchronous path this guide names is not available until they are (Jobs).
Related sheets
- Sheet
- 06 of 20
- Detail
- A
- Updated
- 2026-11-16