DEVELOPER DOCS

Build with Malika™

RESTful API, real-time webhooks, and SDKs for Python, Node.js, and Go. Integrate meeting intelligence into your workflow in minutes.

Authentication
All API requests require a Bearer token. Generate API keys from your Malika™ dashboard under Settings → API Keys. Enterprise (BYOK) users configure their own OpenAI or Anthropic keys separately.
Authentication header
// Every request requires your API key
Authorization: Bearer mk_live_your_api_key_here
Content-Type: application/json
API Endpoints
Base URL: https://api.malikaflow.com/v1
POST
/sessions

Create a new recording session. Attach to a client and calendar event.

POST
/sessions/{id}/audio

Upload audio file (WAV, MP3, M4A, OGG). Max 4 hours. Multipart upload.

GET
/sessions/{id}

Retrieve session with transcript, summary, actions, and tone data.

GET
/sessions/{id}/transcript

Full transcript with word-level timestamps and speaker labels.

POST
/sessions/{id}/analyze

Run one or more engines. Specify: summary, actions, tone, memory, speakers.

GET
/clients/{id}/memory

Retrieve accumulated client memory across all sessions.

POST
/search

Cross-session semantic search. Query across all sessions with source refs.

DELETE
/sessions/{id}/audio

Manually delete audio before retention window. Irreversible.

Create session & analyzePOST
# Upload audio and run all engines
curl -X POST https://api.malikaflow.com/v1/sessions \
  -H "Authorization: Bearer mk_live_xxx" \
  -F "[email protected]" \
  -F "client_id=cl_abc123" \
  -F "engines=summary,actions,tone,memory,speakers"

# Response (202 Accepted)
{
  "session_id": "ses_7f3k2m",
  "status": "processing",
  "estimated_seconds": 45,
  "credits_used": 8,
  "webhook_url": "configured"
}
Webhooks
Real-time notifications when sessions are processed. Configure webhook URLs in your dashboard. All payloads signed with HMAC-SHA256.
1

Audio uploaded

2

Transcription complete

3

Engines processed

4

Webhook fired

Webhook payloadWEBHOOK
{
  "event": "session.analyzed",
  "session_id": "ses_7f3k2m",
  "client_id": "cl_abc123",
  "engines_completed": ["summary", "actions", "tone", "memory", "speakers"],
  "credits_used": 8,
  "results_url": "https://api.malikaflow.com/v1/sessions/se>

  
SDKs
Official client libraries. All open-source on GitHub.

Python

3.8+ · async support

pip install malika

Go

1.21+ · context support

go get malikaflow.com/sdk

Node.js

18+ · TypeScript types

npm i @malika/sdk
Python SDK example
from malika import Malika

client = Malika(api_key="mk_live_xxx")

# Upload and analyze in one call
session = client.sessions.create(
    audio="meeting.mp3",
    client_id="cl_abc123",
    engines=["summary", "actions", "tone", "memory"]
)

# Wait for processing
result = session.wait()

print(result.summary.text)
print(result.actions)           # List of action items
print(result.tone.overall)       # "engaged", "tense", "neutral"
print(result.memory.updates)     # Client memory deltas
Rate Limits & Credits
Rate limits are per API key. Credits are consumed per engine run, not per API call.
PlanRequests/minConcurrent uploadsCredits/monthMax audio
Premium603500 IC4h
Enterprise (BYOK)30010Unlimited4h
EngineCreditsAvg. processing
Deep Summary2 IC~15s
Action Extraction1 IC~8s
Tone Map2 IC~20s
Client Memory1 IC~10s
Cross-Session Search5 IC~5s
Speaker Recap2 IC~12s
Security & Privacy
Designed for professionals handling sensitive conversations.
ENCRYPTION

AES-256 at rest. TLS 1.3 in transit. Audio encrypted before storage.

RETENTION

Premium: 7-day audio retention. Enterprise: configurable 1-30 days. Transcripts persist until deleted.

COMPLIANCE

GDPR / AVG compliant. Data processed in EU (Frankfurt). DPA available on request.

BYOK

Enterprise: bring your own OpenAI or Anthropic key. Your data, your models, your control.

Ready to integrate?

Start your 14-day trial. API keys are available immediately in your dashboard.

Start Free Trial