API reference
API reference overview
Base URL, authentication, common headers, error format and versioning.
Base URL
https://blaze.e2enetworks.net/v1Every endpoint below is relative to this base URL.
Authentication
All requests require a Bearer token:
Authorization: Bearer blz_...Only Authorization: Bearer is accepted — the Anthropic-style x-api-key
header is not read. See Authentication & API keys.
Common headers
| Header | Direction | Purpose |
|---|---|---|
Authorization | Request | Bearer <key> |
Content-Type | Request | application/json (or multipart/form-data for file uploads) |
x-session-id | Request | Keep a conversation on a warm cache — see Session affinity |
x-model | Request | Routing hint; the body's model wins when present |
X-Request-Id | Response | Request ID on inference responses. If you send your own, it is echoed back. Quote it in support requests |
Error format
BlazeInfer returns errors as JSON in the OpenAI shape. The one exception is a 429
from a credit rule, which may be plain text.
- Authentication errors (
401) have amessageand atype:api key requiredwhen no key is sent, andinvalid api keywhen the key is wrong, expired, revoked or blocked. For example{"error":{"message":"api key required","type":"invalid_request_error"}}. - Model errors also have a
code. An unknown or unavailable model returns{"error":{"message":"The requested model is not currently available.","type":"not_found","code":404}}. - Requests the model rejects (
400) get a generic message instead of the original one:{"error":{"message":"The upstream provider rejected the request.","type":"invalid_request","code":400}}.
See Errors & retries for the status codes and what to retry.
Versioning
The API is versioned by the /v1 path segment.
Machine-readable spec
An OpenAPI description of the supported endpoints is published at
/openapi.yaml.
Endpoints
| Endpoint | Status |
|---|---|
| Create chat completion | Supported |
| List models | Supported |
| Create response | |
| Create message (Anthropic) |