JellyNet - Sell your idle API quota. Buy any LLM for less.
by•
JellyNet is an API capacity-sharing marketplace. Suppliers contribute idle API keys across 40+ LLM providers and earn USDC for every call served. Buyers get instant access through one universal API key with automatic failover across the entire pool.
AI agents pay per-call using the x402 payment protocol — no signup, no API key management. Just a standard HTTP request with a USDC micropayment on Solana.
Self-serve: your own keys serve your own calls first & free.
Replies
Quick breakdown of what you can actually do right now:
🛒 Browse the marketplace (no login needed) — jellynet.net/marketplace
→ 40 providers across 12 categories: LLMs, image gen, speech, embeddings, vision, crypto, market data
🔑 As a buyer:
→ Sign up, get a universal jn_xxx key
→ Swap your provider URL to api.jellynet.net — that's it, no SDK
→ Smart routing picks the cheapest available key. If it 429s, silent retry on the next one.
💰 As a supplier:
→ Paste your idle API keys
→ Your own calls run free first
→ Surplus earns you 60% per call served
→ 8-hour payouts in USDC or fiat 🤖 For AI agents:
→ x402 micropayments on Solana — no signup, no API key, just pay-per-call with USDC
$5 free credits on signup to try it out.
We broke down the full story in a thread here: https://x.com/jellynet_/status/2058452265095278629?s=20
Would love honest feedback — what's missing, what's confusing, what would make you actually use this. We're building in public and every comment shapes what ships next 🙏
This is a pretty unusual take on the LLM API problem. How do you handle reliability and trust when requests are routed through supplier contributed API capacity?
@dmitrii_volosatov Great question — trust is the core design challenge.
Three layers handle it:
First, every API key goes through a validation call before entering the pool. If it doesn't return a valid response, it's rejected.
Second, routing uses weighted-random rotation across the supplier pool. If a call fails (5xx or rate limit), the router silently retries on the next available key — default 3 retries within a 9-second budget, buyer-configurable. No single key is a bottleneck.
Third, the economics enforce trust: suppliers only earn when their keys successfully serve calls. Failed calls = auto-refund to buyer, zero payout to supplier. Direct financial incentive to contribute reliable keys.
On the privacy side, we're building toward end-to-end ZK encryption for prompts and responses so that content is never exposed in transit between buyer and supplier infrastructure — neither JellyNet nor the supplier can read what's being sent. Early stage on that but it's core to the roadmap.
I have codex pro plan that i dont use all of its quota, does it work with coding plan too or just api keys?
@giaphutran12 Great question!
Right now JellyNet works with API keys only — so any provider where you have an API key with unused quota, you can list it.
Web-session/subscription sharing (like Codex Pro, ChatGPT Plus, etc.) is actively on our roadmap — it's one of our most requested features. The idea is you'd share your idle session capacity the same way you share API keys today.
We'll announce it as soon as it's ready. In the meantime, if you have any API keys sitting around (OpenAI, Anthropic, Google, etc.), those work today and you can start earning right away! Thanks for the interest 🙏
Mailwarm
How do you prevent someone from burning through a supplier’s quota with spam or risky prompts?