Requesty
OpenClaw

Run OpenClaw on Claude, GPT and Gemini, with the provider built in

Requesty ships with OpenClaw as a bundled provider. Onboard with a Requesty key and OpenClaw discovers the models that key can use, with live prices and context windows. No plugin to install and no JSON to write.

Read the setup guide
bundled with OpenClawREQUESTY_API_KEYmodels discovered from your key
terminal
bundled provider
openclaw onboard --auth-choice requesty-api-key

# or set the key yourself
export REQUESTY_API_KEY="your_requesty_api_key"

openclaw models list --provider requesty
Setup

Set up OpenClaw in three steps

The provider is bundled and updates together with OpenClaw, so there is nothing separate to install. If the gateway runs as a daemon, put REQUESTY_API_KEY in ~/.openclaw/.env so the background process can read it.

  1. 01
    Create a key

    Make an API key on the Requesty API Keys page. The models OpenClaw lists follow what that key is approved to use.

  2. 02
    Run onboarding

    Run openclaw onboard --auth-choice requesty-api-key and paste the key, or set REQUESTY_API_KEY. Onboarding keeps the model entries you already have.

  3. 03
    Pick a model

    The default is requesty/claude-sonnet-5-5. Run /models requesty or openclaw models list --provider requesty to see everything your key can use.

openclaw config
json5
{
  env: { vars: { REQUESTY_API_KEY: "your_requesty_api_key" } },
  agents: {
    defaults: {
      model: { primary: "requesty/claude-sonnet-5-5" },
    },
  },
}
OpenClaw setup in the docsCreate an API key
Models

Models discovered from your key

Once a key is set, OpenClaw asks Requesty for the managed models that key can route to, with live prices, context windows and capability flags. Managed models use bare names, and Requesty picks the hosting provider.

To pin a specific host, add its ID under models.providers.requesty.models and OpenClaw lists it next to the discovered models.

Browse the Model LibraryRouting policies
/models requesty
via requesty
Model IDs that work here
requesty/claude-sonnet-5-5default
requesty/gpt-6-lunamanaged
requesty/gemini-3.8-flashmanaged
requesty/kimi-k3managed
requesty/bedrock/claude-sonnet-4-6pinned host, added in config
99.25%

eventual success on a routing policy in April 2026, against 85.01% for callers pinned to one provider.

Open data note

65.8%

of provider failures on the gateway in April 2026 were 429s. Throttling, not downtime, is what stops an agent mid-task.

Open data note

200

requests a day on free models, with no credit card, to try OpenClaw on Requesty first.

Details

Spend controls

One bill and one log for every channel

OpenClaw answers on WhatsApp, Telegram, Slack and more. Through Requesty, every reply lands in one dashboard attributed to OpenClaw, against one key with its own monthly cap.

  • Attributed to OpenClawRequests carry X-Title: OpenClaw, so usage is split from your other tools
  • A cap on the keyAn always-on assistant stops at a monthly limit you set
  • Approved models onlyDiscovery lists only what the key is allowed to use
Spend limits in the docsSee spend management
keys / sk-agent-eval
capped
monthly cap$500 / $500
requests refused until the cap is raised
80% of the cap
email to the key owner
14 Apr
95% of the cap
webhook to #ai-spend
16 Apr
cap reached
key paused, other keys unaffected
17 Apr

The loop stopped at the ceiling, not at the invoice.

Questions

OpenClaw with Requesty, answered

How do I set up OpenClaw with Requesty?

Requesty is bundled with OpenClaw. Run openclaw onboard --auth-choice requesty-api-key and paste your Requesty API key, or export REQUESTY_API_KEY. Then run openclaw models list --provider requesty to confirm the models are available.

Which models can OpenClaw use through Requesty?

The managed models your key can route to, discovered automatically with live prices and context windows, such as claude-sonnet-5-5, gpt-6-luna and gemini-3.8-flash. The default is requesty/claude-sonnet-5-5. Add any other Model Library ID under models.providers.requesty.models to pin a specific host.

Where does OpenClaw read the Requesty key from?

From REQUESTY_API_KEY, which onboarding sets for you. When the OpenClaw gateway runs as a daemon, make sure the key is available to that process, for example in ~/.openclaw/.env.

How do I change the default model in OpenClaw?

Set agents.defaults.model.primary in your OpenClaw config to a Requesty model reference, such as requesty/gpt-6-luna. Run /models requesty to see which references your key can use.

Do I need to install a plugin for Requesty in OpenClaw?

No. Requesty ships with OpenClaw as a bundled provider plugin and updates with OpenClaw. Older OpenClaw versions can still connect through a custom provider entry in openclaw.json, as our docs show.

How much does it cost to run OpenClaw on Requesty?

OpenClaw is open source, so you pay for model tokens: the provider price plus 5% from a prepaid Requesty balance. Discovery shows each model with its live price, and a monthly cap on the key keeps an always-on assistant within budget.

Other agents

Guides for the other agents you run

The same key and the same dashboard work across every harness, so spend and logs stay in one place.

Claude CodeTwo environment variablesOpenCodeBuilt-in provider, run /connectHermes AgentNative plugin in the catalogPiOne extension, models load at startupKilo CodeBuilt-in provider in SettingsGitHub CopilotBYOK Custom Endpoint in VS Code
All coding agents on Requesty

Run OpenClaw on Requesty

600+ models, a spend cap on every key, and every session on the record.

Speak to founders