Free public beta — no key required · placeholders like ---------- are accepted
DGDG AI APIFree AI Hub
checkingModelsPlaygroundDocsVerifyTry live
DG AI API · Live

DG AI API
Free models.

One OpenAI-compatible endpoint for chat, code, vision, image and voice. Every model free. No signup, no key management — just build.

98.7% Uptime· 90-day rollingFree models · $0 forever
Browse modelsOpen playground
18models live
3providers
98.7%uptime
$0forever
curl — copy / paste ready
# chat — no real key needed. Dashes are accepted.
curl -X POST LOCATION_PLACEHOLDER/v1/chat \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer ----------" \
  -d '{"model":"gpt-oss","messages":[{"role":"user","content":"Hello!"}]}'
Tieranon
Endpoint/v1/chat
Authoptional
Auth is optional. Send Authorization: Bearer ---------- (any number of dashes), X-API-Key: ----------, or nothing — all accepted as anonymous. A real token is enforced only if SECRET_KEY is configured.
Powered byCloudflare Workers AIfreeAI gatewaysSurfSenseNVIDIA NIM
01 — Models

Model page

Every card is live. Click any card to load it straight into the playground below — chat, image and voice switch automatically.

18 shown
Search
02 — Playground

Playground page

Talk to any model. Streaming, history, system prompts and one-click placeholder keys. Image models return pictures, voice models return audio.

Request

Runs against your current Worker origin. Keys are optional.

Dashes, empty, or a real key — all accepted. Sent as Authorization: Bearer ...
0.7
Stream response (SSE)
Playground ready. Pick a model on the left, type below, press Enter. Streaming supported. Placeholders like ---------- work.
idle
03 — Docs

Docs

Minimal integration guide. Full reference lives at /docs.

Full reference

Chat (simple)

Returns { ok, reply, model }. Fastest path for custom apps.

POST /v1/chat
{"model":"gpt-oss", "messages":[{"role":"user", "content":"Hello!"}]}

Chat (OpenAI)

Drop-in completions. Add stream:true for SSE tokens.

POST /v1/chat/completions
{"model":"claude-sonnet-3.5", "messages":[...], "stream":true}

Pinned + media

Lock a model, render images, synthesize voice.

POST /v1/chat/:id · /v1/image
POST /v1/chat/glm POST /v1/image {"prompt":"..."} POST /v1/chat/dg-voice-1

Auth + limits

Anonymous by default. Verify for higher quotas.

anon 8/10s · verified 16/10s
Authorization: Bearer ---------- X-API-Key: ---------- Verify at /verify for images
MethodPathNotes
POST/v1/chatSimple chat, history supported
POST/v1/chat/completionsOpenAI format, streaming
POST/v1/chat/:idPinned model, voice takes { input }
POST/v1/imageFlux PNG, verified only
GET/v1/models · /health · /v1/tierDiscovery + status