One call, then you are calling an LLM

Ask us what is answering right now. The reply carries everything the next call needs.

Base URL, model ID, whether a key is needed, what a real probe proved. Ranked, so models[0] is the pick and the rest are your fallbacks — which you will want, because keyless quotas are per-IP and the one that answered us can refuse you.

1 Ask what works

GET /api/llm/up no key, no signup, no rate limit

curl -s https://stillworks.supercapybara.com/api/llm/up

every model carries modelopenai_base_urlauthlatency_msansweredlast_okproved[]claimed_unproved[]failed[]

?feature=tools,vision narrows it to what a real call proved

2 Call what came back

nothing hardcoded but our URL

python

import json, urllib.request
from openai import OpenAI

shelf = json.load(urllib.request.urlopen(
    "https://stillworks.supercapybara.com/api/llm/up"))

for endpoint in shelf["models"]:            # already ranked, best first
    client = OpenAI(base_url=endpoint["openai_base_url"],
                    api_key="not-needed")
    try:
        reply = client.chat.completions.create(
            model=endpoint["model"],
            messages=[{"role": "user", "content": "hello"}],
        )
    except Exception:
        continue                           # your IP hit its quota — next
    print(endpoint["model"], reply.choices[0].message.content)
    break

node

import OpenAI from "openai";

const shelf = await fetch("https://stillworks.supercapybara.com/api/llm/up")
  .then((response) => response.json());

for (const endpoint of shelf.models) {      // already ranked, best first
  const client = new OpenAI({
    baseURL: endpoint.openai_base_url,
    apiKey: "not-needed",
  });
  try {
    const reply = await client.chat.completions.create({
      model: endpoint.model,
      messages: [{ role: "user", content: "hello" }],
    });
    console.log(endpoint.model, reply.choices[0].message.content);
    break;
  } catch { continue; }                    // your IP hit its quota — next
}

curl

# whatever your JSON tool is, the two fields you need are the same
read BASE MODEL < <(curl -s https://stillworks.supercapybara.com/api/llm/up \
  | jq -r '.models[0] | "\(.openai_base_url) \(.model)"')

curl -s "$BASE/chat/completions" \
  -H 'Content-Type: application/json' \
  -d "{\"model\":\"$MODEL\",
       \"messages\":[{\"role\":\"user\",\"content\":\"hello\"}]}"

Walk the list until one answers. That loop is the whole point of the shelf.

KeyAny Not needed Required

Proved Tools Schema JSON Vision

8 of 100 models · proved vision

No key needed first, then by how much of the last week each one answered, then by speed. None of it is a promise about the next minute — that is what the timestamps are for. Click any row for its snippet.

we called it and it worked we called it and it failed provider claims it, our call never proved it never probedWhat Tools, Schema, JSON and Vision mean

Expand	Model▲	Key	Tools	Schema	JSON	Vision	Latency▲	Answered▲	Verified▲
›	Lorbus/Qwen3.6-27B-int4-AutoRound uncloseai · hermes.ai.unturf.com/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	1674 ms	101/151	34 h ago
curl `curl https://hermes.ai.unturf.com/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Lorbus/Qwen3.6-27B-int4-AutoRound","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://hermes.ai.unturf.com/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Lorbus/Qwen3.6-27B-int4-AutoRound` python `from openai import OpenAI client = OpenAI( base_url="https://hermes.ai.unturf.com/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Lorbus/Qwen3.6-27B-int4-AutoRound", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://hermes.ai.unturf.com/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Lorbus/Qwen3.6-27B-int4-AutoRound", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://hermes.ai.unturf.com/v1 Model ID Lorbus/Qwen3.6-27B-int4-AutoRound Auth none — no Authorization header sent Answered answered 101 of the 151 checks we made in the last week Last ok 34 h ago
›	Qwen3.8-27B ovh-anonymous · oai.endpoints.kepler.ai.cloud.ovh.net/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	642 ms	4/10	30 h ago
curl `curl https://oai.endpoints.kepler.ai.cloud.ovh.net/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Qwen3.8-27B","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Qwen3.8-27B` python `from openai import OpenAI client = OpenAI( base_url="https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Qwen3.8-27B", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Qwen3.8-27B", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 Model ID Qwen3.8-27B Auth none — no Authorization header sent Answered answered 4 of the 10 checks we made in the last week Last ok 30 h ago
›	RedHatAI/Qwen3.8-27B-INT4 uncloseai · hermes.ai.unturf.com/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	783 ms	4/17	25 h ago
curl `curl https://hermes.ai.unturf.com/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"RedHatAI/Qwen3.8-27B-INT4","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://hermes.ai.unturf.com/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=RedHatAI/Qwen3.8-27B-INT4` python `from openai import OpenAI client = OpenAI( base_url="https://hermes.ai.unturf.com/v1", api_key="not-needed", ) print(client.chat.completions.create( model="RedHatAI/Qwen3.8-27B-INT4", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://hermes.ai.unturf.com/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "RedHatAI/Qwen3.8-27B-INT4", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://hermes.ai.unturf.com/v1 Model ID RedHatAI/Qwen3.8-27B-INT4 Auth none — no Authorization header sent Answered answered 4 of the 17 checks we made in the last week Last ok 25 h ago
›	Qwen2.5-VL-72B-Instruct ovh-anonymous · oai.endpoints.kepler.ai.cloud.ovh.net/v1	not needed	Tools: we called it and it failed	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	769 ms	2/10	2 days ago
curl `curl https://oai.endpoints.kepler.ai.cloud.ovh.net/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Qwen2.5-VL-72B-Instruct","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Qwen2.5-VL-72B-Instruct` python `from openai import OpenAI client = OpenAI( base_url="https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Qwen2.5-VL-72B-Instruct", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Qwen2.5-VL-72B-Instruct", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 Model ID Qwen2.5-VL-72B-Instruct Auth none — no Authorization header sent Answered answered 2 of the 10 checks we made in the last week Last ok 2 days ago
›	Qwen3.6-27B ovh-anonymous · oai.endpoints.kepler.ai.cloud.ovh.net/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	729 ms	2/11	4 days ago
curl `curl https://oai.endpoints.kepler.ai.cloud.ovh.net/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Qwen3.6-27B","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Qwen3.6-27B` python `from openai import OpenAI client = OpenAI( base_url="https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Qwen3.6-27B", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Qwen3.6-27B", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 Model ID Qwen3.6-27B Auth none — no Authorization header sent Answered answered 2 of the 11 checks we made in the last week Last ok 4 days ago
›	Qwen3.5-397B-A17B ovh-anonymous · oai.endpoints.kepler.ai.cloud.ovh.net/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: never probed	Vision: we called it and it worked	468 ms	0/10	7 days ago
curl `curl https://oai.endpoints.kepler.ai.cloud.ovh.net/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Qwen3.5-397B-A17B","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Qwen3.5-397B-A17B` python `from openai import OpenAI client = OpenAI( base_url="https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Qwen3.5-397B-A17B", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Qwen3.5-397B-A17B", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 Model ID Qwen3.5-397B-A17B Auth none — no Authorization header sent Answered answered 0 of the 10 checks we made in the last week Last ok 7 days ago
›	Mistral-Small-3.2-24B-Instruct-2506 ovh-anonymous · oai.endpoints.kepler.ai.cloud.ovh.net/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	670 ms	0/11	12 days ago
curl `curl https://oai.endpoints.kepler.ai.cloud.ovh.net/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Mistral-Small-3.2-24B-Instruct-2506","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Mistral-Small-3.2-24B-Instruct-2506` python `from openai import OpenAI client = OpenAI( base_url="https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Mistral-Small-3.2-24B-Instruct-2506", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Mistral-Small-3.2-24B-Instruct-2506", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 Model ID Mistral-Small-3.2-24B-Instruct-2506 Auth none — no Authorization header sent Answered answered 0 of the 11 checks we made in the last week Last ok 12 days ago
›	Qwen3.5-9B ovh-anonymous · oai.endpoints.kepler.ai.cloud.ovh.net/v1	not needed	Tools: we called it and it worked	Schema: we called it and it worked	JSON: we called it and it worked	Vision: we called it and it worked	1162 ms	0/11	5 days ago
curl `curl https://oai.endpoints.kepler.ai.cloud.ovh.net/v1/chat/completions \ -H 'Content-Type: application/json' \ -d '{"model":"Qwen3.5-9B","messages":[{"role":"user","content":"hello"}]}'` .env `OPENAI_BASE_URL=https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 OPENAI_API_KEY=not-needed OPENAI_MODEL=Qwen3.5-9B` python `from openai import OpenAI client = OpenAI( base_url="https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", api_key="not-needed", ) print(client.chat.completions.create( model="Qwen3.5-9B", messages=[{"role": "user", "content": "hello"}], ).choices[0].message.content)` node `import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1", apiKey: "not-needed", }); const reply = await client.chat.completions.create({ model: "Qwen3.5-9B", messages: [{ role: "user", content: "hello" }], }); console.log(reply.choices[0].message.content);` Base URL https://oai.endpoints.kepler.ai.cloud.ovh.net/v1 Model ID Qwen3.5-9B Auth none — no Authorization header sent Answered answered 0 of the 11 checks we made in the last week Last ok 5 days ago

Checked and not listed (8) — endpoints another catalogue calls free, and what they answered here

Cross-checked against OmniRoute's routing registry on 2026-08-11: 226 providers, 18 of them typed as needing no key. Calling all 18 from this server produced one new claim on the shelf above and no new working endpoint.

Hack Club AI https://ai.hackclub.com/proxy/v1: Listing answers a bare GET with 200 and a full model catalogue; the chat call answers 401 Authentication required. A keyless /v1/models is not a keyless endpoint, and this is the shape most often mistaken for one. Separately scoped to Hack Club members by its own terms.
Naga https://api.naga.ac/v1: Same shape: listing 200, chat 401 authentication_required, with the body directing you to sign up. A free tier behind a signup is the keyed shelf's subject, not this one, and we hold no key.
Pollinations (gen host) https://gen.pollinations.ai/v1: 401 UNAUTHORIZED with no header. Worth stating plainly because the catalogue routes here on the grounds that the older text.pollinations.ai host was retired -- and that older host answered our keyless chat call with 200 on the same day. The retired endpoint is the one still serving anonymously, so that is the one on the shelf.
g4f.space https://g4f.space/api/{pollinations,groq,gemini,nvidia,ollama}/v1: 402 insufficient_credits: "No cake credits. Bake proof-of-work cakes ... to earn anonymous usage." Anonymous, but not free of charge -- the price is compute rather than a key. Nothing an OpenAI client can pay, so it cannot be published as an endpoint that works right now.
MiMo (Xiaomi) https://api.xiaomimimo.com/v1: 401 invalid_key on both the listing and the chat call.
TheOldLLM https://theoldllm.vercel.app/api/chatgpt: 403 at the edge. The same catalogue's own free-tier research lists this provider as having no current free tier, so the registry entry and the documentation disagree; the call settles it.
Web-UI scrapers duckduckgo-web, felo-web, qwen-web, t3-web, muse-spark-web, veoaifree-web, agy: Not probed, and not a measurement question. These drive somebody's chat website through an adapter; the source catalogue's own terms-of-service table marks every one of them avoid for proxied or non-personal use. An endpoint we would have to breach terms to call is not an endpoint we can tell you to call.
Vendor agent CLIs auggie, devin-cli-agentic, kiro, amazon-q, opencode: Credential-free only in the sense that an installed CLI already holds an OAuth session. There is no URL to hand an OpenAI client, so there is nothing here for this shelf to check.

What a row means

We send a real chat completion to the endpoint on a schedule and publish what came back — latency, how many of the recent checks answered, and when the last one did.

Keyless rows carried no Authorization header at all. Rows marked key carried our own free-tier key, which proves the endpoint answered our account — not what your signup gets.

What we cannot tell you

Keyless quotas are usually per-IP. An endpoint verified from this server can still answer 402 or 429 from yours.

These are other people's free services. Any of them can add a key requirement or disappear between two probes.

The table is this API

Every filter here is a query parameter there: ?feature=tools,vision, ?key=none, ?q=qwen. Same ranking, same rows.

Filters live in the URL, so a filtered shelf is a link you can send someone.

What Tools, Schema, JSON and Vision mean, and what each cell state means

The columns

The call behind any one verdict is on that endpoint's own page — follow the provider link in a row and read The call behind each verdict.

Tools: We sent a tools array and the model came back with a well-formed tool_calls entry an agent could dispatch. A 200 with no tool call in it is not tool support, and is recorded as a no.
Schema: The strict structured-output mode: response_format: {"type": "json_schema", …}. The reply parsed and matched the schema we sent.
JSON: The weaker mode: response_format: {"type": "json_object"}. Valid JSON, with no shape enforced. The two come apart in both directions — OVH's Mistral-7B-Instruct-v0.3 does the strict form and emits invalid JSON for the weak one; llm7's meta-Llama-3.1-8B-Instruct-Turbo does json_object while json_schema answers 405.
Vision: A real image went in inline and a correct answer about it came back. We ask about something only a model that looked could know.
Draws: On an endpoint's own page: bytes came back with an image content type, and they really are an image. Not a 200 taken on trust.

The cell states

yes: A real call proved it, at the time the endpoint page prints beside that call. The dot's tooltip says the same thing where there is a pointer to hover with.
no: A clean negative: the parameter was refused, or the answer did not conform.
no, claimed: The negative contradicts the provider's own published claim — their listing advertises the feature and our call could not get it out of the model. This is the most interesting cell on the site and the reason the shelf exists.
never probed: No verdict. Either nobody has asked yet, or the check ran and got no usable answer back — Pollinations sits here because its anonymous pool answered 402 to every feature probe. It is a question, never a no.

The same columns, in the API

curl 'https://stillworks.supercapybara.com/api/llm/up?feature=tools,json_schema'

The names are the API's spelling of the columns: tools, json_schema, json_object, vision, image_out. Several are and, never or — a row comes back only if a real call proved every feature named. Models we never probed for one of them are left out rather than optimistically included, so a filtered shelf is shorter than a hopeful one.