wehub-resource-sync
17 lines
893 B
TOML
17 lines
893 B
TOML
name = "routing.run"
|
|
env = ["ROUTING_RUN_API_KEY"]
|
|
npm = "@ai-sdk/openai-compatible"
|
|
api = "https://api.routing.run/v1"
|
|
# Per-model `cost` and `limit` are taken from the live catalog endpoint
|
|
# GET https://api.routing.run/v1/models (accessed 2026-07-09), which returns
|
|
# `pricing` (prompt/completion, per-token) and `limit` (context/output) for
|
|
# every served model. That endpoint exposes no cache-tier pricing, so
|
|
# `cost.cache_read` is intentionally omitted. Public pricing page:
|
|
# https://docs.routing.run/api-reference/models
|
|
# Reasoning HTTP format (accessed 2026-07-09):
|
|
# POST /v1/chat/completions returns `message.reasoning_content` for
|
|
# reasoning-capable models. The request body exposes no reasoning toggle,
|
|
# effort, or token-budget control, so `reasoning_options = []`.
|
|
# Source: https://docs.routing.run/api-reference/openai
|
|
doc = "https://docs.routing.run/api-reference/models"
|