Launching today
Synap402
OpenAI-compatible API for open models at near-cost prices
1 follower
OpenAI-compatible API for open models at near-cost prices
1 follower
One OpenAI-compatible endpoint for open-weight models (Qwen3 Coder, gpt-oss-20b, Gemma 4, Qwen3-VL) priced near cost: gpt-oss-20b is $0.017 in / $0.085 out per 1M tokens. Free tier: 100K tokens/day. Agents can pay in USDC over x402, no account.
Free Options
Launch Team
AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted
Hi Product Hunt! We built Synap402 because most of what agents and coding tools send is small, repetitive work that does not need a frontier model.
Synap402 is one OpenAI-compatible endpoint (https://synap402.com/v1) for open-weight models. Change base_url and key in any OpenAI SDK, Claude Code, Codex, Cline or Aider.
Live prices per 1M tokens (in/out): gpt-oss-20b $0.017/$0.085, Qwen3 Coder 30B $0.0675/$0.27, Gemma 4 26B $0.04/$0.21. Batch is half.
Two things we care about: privacy (we never train on your prompts, and we don't keep the text of your API prompts or replies), and x402 payments, so an agent with a USDC wallet can top up on its own with no account.
There is a free tier (100K tokens a day on Qwen3 Coder, no card). We would love feedback on which models you want next.