Audaxic Memory is a 100% mem0-compatible memory layer running on Cloudflare edge. Same V3 pipeline, same API. Flat pricing starting at $0/mo.
Everything you need for AI agent memory, running on the edge.
8-phase additive extraction, deduplication, entity boosting — identical to mem0's latest pipeline.
Hybrid search with sigmoid BM25 fusion and entity boost. Returns scored results with explanations.
Runs on Cloudflare Workers with zero native deps. Regex entity extraction, D1 storage, Vectorize.
Organization-scoped namespaces, per-org API keys, isolated vector search. Built for SaaS.
Deployed on Cloudflare's global edge network. Sub-100ms latency worldwide. No cold starts.
Same request/response shapes as mem0. Change base_url, that's it. JS, Python, and REST SDKs included.
Install, import, done. Works with your existing mem0 code.
import { Memory } from "@audaxic/memory";
const m = new Memory({
apiKey: "ax_...",
baseUrl: "https://api.audaxic.com",
});
// Add memories
await m.add(
[{ role: "user", content: "I like cats" }],
{ userId: "alice" }
);
// Search
const results = await m.search(
"what does alice like?",
{ userId: "alice" }
);from audaxic import Memory
m = Memory(api_key="ax_...")
# Add memories
m.add("I like cats", user_id="alice")
# Search
results = m.search(
"what does alice like?",
user_id="alice"
)No per-token fees. No surprises. Scale as you grow.
1,000 memories · 10 req/min
10,000 memories · 60 req/min
1,000,000 memories · 1,000 req/min
100,000 memories · 300 req/min
Same API. Better infrastructure. Lower cost.
| Feature | Mem0 | Audaxic |
|---|---|---|
| Memory operations | Per-token pricing | Flat monthly |
| Infrastructure | Your cloud account | Fully managed |
| Setup time | ~30 minutes | < 5 minutes |
| API compatibility | 100% | 100% drop-in |
| Entity extraction | Python (spaCy) | Workers-native (zero deps) |
| Hosting | Self-hosted only | Edge (Cloudflare) |
| Pricing | $19–$249+/mo | $0–$125/mo |
Audaxic Memory is a drop-in replacement for mem0 — a memory layer for AI agents. It stores, retrieves, and manages facts learned from conversations, giving your AI persistent context across sessions.
Same API, same V3 pipeline, but running entirely on Cloudflare edge (Workers + D1 + Vectorize). No cloud accounts to manage, no per-token billing, flat monthly pricing at ~50% less.
Yes. Install the SDK, change your base_url and api_key, and everything works. Same request/response shapes, same error codes, same behavior.
Your data stays on Cloudflare's edge network. We never train on your data. Each organization has isolated storage with namespace-scoped vector search.
Yes. The API runs on Cloudflare Workers — deploy to your own Cloudflare account with wrangler. Full source code available.
We use gpt-oss-20b (via Lightning.ai) for entity extraction and Qwen3-Embedding-8B (via DeepInfra) for embeddings. Both are production-grade, cost-effective models.
Start free. No credit card required. Deploy in minutes.