Frasberg — Verified LLM Provider

Call Luchii in 60 seconds.

Frasberg is an OpenAI-compatible provider with Bearer authentication and SSE streaming. Point any OpenAI SDK, Vercel AI SDK, or LangChain at https://api.frasberg.com/v1 and it just works.

Bearer auth

Authorization: Bearer frb_live_…

SSE streaming

stream: true — chat.completion.chunk

Verified provider

/.well-known discovery manifests

1. Get your key

Create a free account and forge a frb_live_ key with the permissions matrix in the Developer Dashboard. Every key ships with 2,500 trial tokens — then top up with credit packs at half the price of other providers (from $5 / 10k tokens), or subscribe for unmetered usage.

2. Make your first call

curl https://api.frasberg.com/v1/chat/completions \
  -H "Authorization: Bearer $FRASBERG_LLM_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "luchii-6-plus",
    "messages": [{"role": "user", "content": "Hello Luchii"}],
    "stream": true
  }'

3. Embeddings

curl https://api.frasberg.com/v1/embeddings \
  -H "Authorization: Bearer $FRASBERG_LLM_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "luchii-6-embed", "input": "Intelligence, harmonized."}'

4. Frasberg Engine — Music, Jobs & Audio Tools

Engine endpoints use your frb_live_ keys with the granular permissions matrix. Jobs run asynchronously: queued → running → completed, then the result URL streams the final asset. All engine keys are rate-limited (default 60 req/min).

POST /api/generate/music

# Submit a music generation job (permission: music_generation)
curl -X POST "$API_URL/api/generate/music" \
  -H "Authorization: Bearer frb_live_xxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{"prompt": "haunting dark noir", "duration": 180, "model": "frasberg-music"}'

# 200 OK
# {
#   "job_id": "mtask_20260912T045052Z_c93c6c52",
#   "status": "queued",
#   "eta_seconds": 15,
#   "model": "frasberg-music",
#   "mood": "minor"
# }

GET /api/jobs/{job_id}

# Poll any engine job (music mtask_* or video task_*)
curl -X GET "$API_URL/api/jobs/$JOB_ID" \
  -H "Authorization: Bearer frb_live_xxxxxxxxxxxxx"

# States: queued -> running -> completed
# On completion:
# {
#   "job_id": "mtask_...",
#   "status": "completed",
#   "result": { "url": "/api/generate/music/task/mtask_.../audio" }
# }

# Stream the final audio with the same key:
curl -o track.wav "$API_URL/api/generate/music/task/$JOB_ID/audio" \
  -H "Authorization: Bearer frb_live_xxxxxxxxxxxxx"

POST /api/audio/tools/enhance

# Enhance any audio source — real DSP: noise gate, presence EQ, limiter
# (permission: audio_native — accepts external URLs or your own track URLs)
curl -X POST "$API_URL/api/audio/tools/enhance" \
  -H "Authorization: Bearer frb_live_xxxxxxxxxxxxx" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com/raw-recording.mp3"}'

# 200 OK
# {
#   "job_id": "atask_20260912T095813Z_e23f6420",
#   "status": "completed",
#   "result": { "url": "/api/audio/tools/enhance/atask_.../audio" },
#   "latency_ms": 84
# }

# Download the processed WAV:
curl -o enhanced.wav "$API_URL/api/audio/tools/enhance/$JOB_ID/audio" \
  -H "Authorization: Bearer frb_live_xxxxxxxxxxxxx"

Engine error codes

401FK-001Invalid API key — the frb_live_ key does not exist or was revoked

403FK-003Missing permission — e.g. music_generation or audio_native not granted on this key

422FK-422Source is not decodable audio / source url could not be fetched

429FK-429Rate limit exceeded — default 60 req/min per key, Retry-After header included

Models

luchii-6-pluschatFlagship frontier reasoning — deepest cognition tier

luchii-6-minichatFast, balanced everyday reasoning

luchii-6-embedembed384-dim text embeddings for search & RAG

luchii-70b / 7b / 1b / 200mchatLegacy tier names — fully supported aliases

Legal & compliance

Partner API Agreement Provider SLA Compliance Packet

Provider verification

Standard discovery documents used by registries, routers and agent frameworks: