Manage your AI provider connections — the OmniRoute-style control panel for the gateway. Keys live only in this browser. Interface pattern from OmniRoute ↗.
① Wired (3) — in the router's catalog: auto picks among them, fallback + free-tier headroom included. Connect, ▶ Test, toggle.
② Onboardable (11) — OpenAI-compatible: paste your key, ▶ Test it, then use model: "<provider>:<model>" from any tool. Direct passthrough to their verified API base.
③ Directory — own APIs (dedicated endpoints, media): listed for reference, not generically onboardable.
Wired providers — live in the gateway today: connect, test, toggle
◣
NVIDIA build.nvidia.com (NIM)
free credits · NIM catalog
checking…
get a free key ↗ · stored only in this browser
☁
Cloudflare Workers AI
10k neurons/day free · edge
checking…
get a free key ↗ · stored only in this browser
Fallback chain — watch it work — preview who's ranked, then run a REAL /v1 request and see who answers
Pick an alias and hit “Preview chain” to see the router's ranked candidates.
Onboardable — bring your key — 11 OpenAI-compatible providers you can add right now: paste a key, ▶ Test, then model: "<provider>:<model>"
🤗 Hugging Face is the meta-router. One HF token reaches 128+ models across many backends (Together, Fireworks, Cerebras, Groq…) — address any as hf:owner/model with a :cheapest or :fastest policy suffix. It's the widest single door here.
H
one HF token → 128+ models across many backends · free monthly credits
bring your key to onboard
use: model: "hf:openai/gpt-oss-120b:cheapest" + your key as Bearer — any OpenAI tool
onboardable meta-router
T
api.together.ai settings > API keys; select models offered in a free tier
bring your key to onboard
use: model: "together:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
F
fireworks.ai account > API keys; new accounts get trial credits
bring your key to onboard
use: model: "fireworks:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
D
deepinfra.com dashboard > API keys; pay-as-you-go pricing
bring your key to onboard
use: model: "deepinfra:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
C
Free key at cloud.cerebras.ai console; generous rate-limited free tier
bring your key to onboard
use: model: "cerebras:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
S
cloud.sambanova.ai > API keys; free developer tier available
bring your key to onboard
use: model: "sambanova:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
N
Key from Token Factory console (tokenfactory.nebius.com); studio keys carry over
bring your key to onboard
use: model: "nebius:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
N
novita.ai dashboard > Key Management; new-user credits offered
bring your key to onboard
use: model: "novita:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
H
Key from Hyperbolic dashboard (hyperbolic.ai); new accounts get a small trial credit
bring your key to onboard
use: model: "hyperbolic:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Serverless model APIs
O
openrouter.ai/keys; many ':free' model variants with daily rate limits
bring your key to onboard
use: model: "openrouter:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable meta-router
B
app.baseten.co API keys; new accounts get free credit
bring your key to onboard
use: model: "baseten:<their-model-id>" + your key as Bearer — any OpenAI tool
onboardable Model serving platforms
Directory — 11 more with their own APIs (dedicated endpoints, media): reference, not generically onboardable
f
1000+ generative media models — image, video, audio, 3D — built for speed
directory — own API, not generically onboardable
Serverless model APIs
per-output on model APIs + hourly GPU for dedicated compute
R
Run thousands of community models with one line of code, pay per second
directory — own API, not generically onboardable
Serverless model APIs
per-second of hardware (T4 $0.000225/s, A100-80GB $0.0014/s)
B
Open-source serving framework plus a managed inference cloud
directory — own API, not generically onboardable
Model serving platforms
BYOC/on-prem/managed options; homepage doesn't state rates
M
Python-native serverless GPUs with sub-second cold starts
directory — own API, not generically onboardable
GPU clouds
per-second GPU/CPU/memory (H100 $0.001097/s, T4 $0.000164/s)
R
Per-millisecond GPU pods and serverless endpoints with sub-200ms cold starts
directory — own API, not generically onboardable
GPU clouds
per-second GPU (billed per-millisecond); serverless scale-to
C
Energy-first AI factory: from power plant to OpenAI-compatible inference API
directory — own API, not generically onboardable
GPU clouds OpenAI-compat
per-hour GPU (on-demand H100 from $3.90/GPU-hr; reserved dis
L
Single-tenant NVIDIA superclusters, 1-Click Clusters, and on-demand instances
directory — own API, not generically onboardable
GPU clouds
per-hour GPU (per-GPU-hour on multi-GPU nodes); volume disco
C
Kubernetes-native AI hyperscaler serving OpenAI, Mistral, and IBM
directory — own API, not generically onboardable
GPU clouds
per-hour GPU + reserved capacity (up to 60% off on-demand);
V
GPU marketplace where prices are set by the market, not the cloud
directory — own API, not generically onboardable
GPU clouds
market-set per-second GPU: on-demand, interruptible (50%+ ch
A
Open models per-token inside the AWS compliance envelope
directory — own API, not generically onboardable
Serverless model APIs
per-token (on-demand) + provisioned throughput
A
Microsoft's model catalog with serverless open-model endpoints
directory — own API, not generically onboardable
Serverless model APIs
per-token (serverless) + managed compute