surp/free — real responses, zero payment, treasury-sponsored.
surp/free gives users genuinely free LLM inference through the same OpenAI-compatible endpoint. There is no x402 payment challenge and no API key. The surp treasury pays the upstream model cost, with strict daily budgets, per-IP limits, cheap-model guardrails, and automatic fallback.
curl -X POST https://surp.ivc.lol/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{"model":"surp/free",
"messages":[{"role":"user","content":"Explain x402 in one sentence"}],
"max_tokens":64,
"stream":false}'
Limits: 20 requests/IP/day · 128 max output tokens · non-streaming · no tool calls · availability is best-effort.
| requests served today | 0 / 500 |
| tokens served today | 0 / 100,000 |
| failed upstream attempts | 0 |
| maximum sponsored model price | $0.1000 / 1M tokens |
| fallback attempts per request | up to 3 live models |
These are real, liquid Surplus marketplace models currently below the configured treasury ceiling. The router tries the cheapest first and falls back when a seller is unavailable.
| # | model | USD/1M | healthy sellers | requests 24h | volume 24h |
|---|---|---|---|---|---|
| 1 | glm-5.2 | $0.0128 | 148 | 49,549 | 19,299,738 |
| 2 | glm-4.7 | $0.0215 | 177 | 7,484 | 1,808,669 |
| 3 | gemma-3-27b-it | $0.0252 | 109 | 1 | 2 |
| 4 | deepseek-v4-flash-0731 | $0.0276 | 42 | 16,625 | 4,489,995 |
| 5 | qwen3-235b-a22b-2507 | $0.0280 | 122 | 0 | 0 |
| 6 | deepseek-v4-flash | $0.0297 | 114 | 281,685 | 73,498,700 |
| 7 | mistral-small-3.2-24b-instruct | $0.0344 | 105 | 48 | 0 |
| 8 | glm-4.5-air | $0.0390 | 36 | 50 | 53 |
| 9 | glm-4.7-flash | $0.0414 | 123 | 63 | 1,205 |
| 10 | e2ee-qwen-2-5-7b-p | $0.0432 | 83 | 1 | 80 |
| 11 | qwen3-5-9b | $0.0500 | 81 | 6 | 2,237 |
| 12 | gemini-2.5-flash-image | $0.0500 | 23 | 0 | 0 |
| 13 | e2ee-gpt-oss-20b-p | $0.0575 | 81 | 0 | 0 |
| 14 | glm-4.6 | $0.0654 | 177 | 54 | 293 |
| 15 | qwen3-235b-a22b-thinking-2507 | $0.0800 | 83 | 70 | 815 |
| 16 | gemma-4-31b-it | $0.0825 | 85 | 160 | 95,584 |
| 17 | glm-4.5 | $0.0840 | 28 | 50 | 1,711 |
| 18 | llama-3.3-70b-instruct | $0.0840 | 104 | 64 | 369 |
| 19 | gpt-5-nano | $0.0900 | 38 | 58 | 1,272 |
| 20 | openai-gpt-oss-120b | $0.0924 | 114 | 54 | 378 |
| 21 | nvidia-nemotron-3-nano-30b-a3b | $0.0937 | 101 | 56 | 25 |
| 22 | gpt-5.6-luna | $0.0980 | 90 | 53,316 | 14,211,585 |
| model | free requests | tokens |
|---|---|---|
| No free requests served today yet. | ||
OmniRoute is an MIT-licensed AI gateway with an unusually careful free-tier catalog. We incorporated the parts that improve honesty and transparency:
Source: OmniRoute release/v3.8.50, curated 2026-07-22, MIT License. Snapshot stored with license attribution in this repository.
| providers represented | 81 |
| permanently free, uncapped providers | 13 (listed, never summed) |
| ToS generally ok | 49 |
| ToS ambiguous | 38 |
| ToS caution | 334 |
| ToS avoid | 93 |
| catalog entries after excluding avoid | 426 |
Because "free" often means free for one developer's account—not free to resell through a public proxy. OmniRoute itself surfaces these ToS risks. surp will only integrate a third-party free provider into public routing when its terms explicitly allow proxying/resale or we have written permission. Until then, the catalog is informational and our public free requests are treasury-sponsored.
related: live system status · cache-aware routing · reward proposal · token-gated access · API docs