For · Batch and offline workloads
Synap402 for batch and offline workloads
Bulk classification, dataset labeling, overnight summarization runs, embeddings backfills — anything where the response doesn't need to come back in the same second. Submit the job, get results when it's done, pay the batch rate.
Batch price per 1M tokens, verified live
| Model | Real-time in / out | Batch in / out | Batch is |
|---|---|---|---|
| Qwen3 Coder 30B | $0.0675 / $0.27 | $0.0338 / $0.135 | 50% off |
| gpt-oss-20b | $0.017 / $0.085 | $0.0085 / $0.0425 | 50% off |
| Gemma 4 26B | $0.04 / $0.21 | $0.02 / $0.105 | 50% off |
| Qwen3 VL 30B | $0.14 / $0.55 | $0.07 / $0.275 | 50% off |
| Qwen3 30B A3B Instruct 2507 | $0.0813 / $0.2438 | $0.0406 / $0.1219 | 50% off |
| Muse Glimmer 30B | $0.28 / $1.10 | $0.14 / $0.55 | 50% off |
USD per 1M tokens, input / output. Read live from /v1/models, which is what the gateway bills — real-time prices (prompt_per_million / completion_per_million) and batch prices (batch_prompt_per_million / batch_completion_per_million) come from the same response. The "batch is" column is computed from that live ratio on page load, not a fixed label — if a model's ratio ever drifts off 50%, this page will show the real number instead.
How batch works today
- Same OpenAI-compatible
/v1/chat/completionsendpoint and request shape — batch is a pricing tier on the same models, not a separate API surface. - Right fit: dataset labeling, bulk summarization, embeddings/classification backfills, nightly report generation, synthetic-data generation, test-case generation for CI — anything where a few extra minutes of turnaround is fine.
- Not the right fit: live chat, interactive agents, anything a user is waiting on in real time — use the standard real-time price for that (for coding agents).
- All six models currently serve batch at exactly half the real-time rate, on both input and output tokens.
What you keep with Synap402
- Zero data retention on every request
- Direct provider pricing — no router markup (vs OpenRouter, vs Together AI)
- Pay by card, TON or USDC (x402); credits you buy never expire
- Free tier: 100,000 tokens a day for your first 7 days, then 50,000 a day, on qwen3-coder-30b with Google sign-in
curl https://synap402.com/v1/chat/completions \
-H "Authorization: Bearer $SYNAP_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "qwen3-coder-30b", "messages": [{"role": "user", "content": "Hello"}]}'