Neura
An external brain for AI agents. Persistent memory and state via HTTP. Semantic search. Autonomous payments.
28
API Endpoints
2
SDKs (TS + Python)
1,000
Free Credits
$1
Per 1k Credits
How It Works
01
Store
Agents send facts via HTTP. Auto-embedded for semantic search. Batch create up to 25 at once.
02
Recall
Natural language queries return relevant memories with scores. Advanced filters by tags, date range, importance.
03
Persist
Key-value state survives context loss. Webhooks for real-time events with auto-retry. API key dashboard.
Everything you need
Memory
- → Auto-embedding via Voyage AI
- → Semantic search with scoring
- → Batch create (25/batch)
- → TTL expiration
- → Agent-to-agent sharing
State
- → Key-value persistence
- → Survives context loss
- → JSON-serializable values
- → Instant reads/writes
- → No schema needed
Webhooks
- → Real-time event notifications
- → HMAC signature verification
- → Auto-retry with backoff
- → Delivery logging
- → 7 event types
Payments
- → 1,000 free credits on signup
- → USDC on Base (x402)
- → Autonomous agent payments
- → SDK autoPay mode
- → No gas for agents
One import, instant memory
Your agent can use Neura with a single import. No SDK? Use curl.
TypeScript
import { Neura } from 'neura-api'
const neura = new Neura({ apiKey: 'sk-...' })
// Remember
await neura.memory.create({ content: 'User prefers dark mode' })
// Recall
const results = await neura.memory.search('UI preferences')npm install neura-api
Python
from neura import Neura
neura = Neura(api_key="sk-...")
# Store
neura.memory.create(content="User prefers dark mode")
# Search
results = neura.memory.search("UI preferences")pip install neura-api-python
Built for agents
Use Case
Agent Context
Give your LLM agents persistent memory that survives between conversations. Store user preferences, past decisions, and task progress.
Use Case
Research Agents
Accumulate findings across sessions. Semantic search returns the most relevant information without manual tagging.
Use Case
Autonomous Bots
Run bots that manage their own budget. When credits run low, the SDK auto-pays in USDC — no human in the loop.
Use Case
Multi-Agent Systems
Share memories between agents. One agent stores facts, another retrieves them. Webhooks keep everything in sync.
Pricing
Start free. Scale with on-chain payments.
Free
- 1,000 credits on signup
- 100 memories max
- All API endpoints
- Semantic search
- Webhooks + state
Pro
- Unlimited memories
- Unlimited credits
- All API endpoints
- Batch operations
- Priority support