This is the catalog for the LLM proxy and voice surface. For an end-to-end walkthrough, see the Tutorial.
Supported providers and proxy base URLs
The proxy is a transparent pipe — it strips the /openai, /anthropic, or /elevenlabs prefix and forwards the rest to the upstream provider. The OpenAI and Anthropic SDKs don't auto-append /v1, so put it in the base URL. For ElevenLabs the version segment (/v1 or /v2) goes in the request path, so the base URL ends at /elevenlabs.
| Provider | Proxy base URL | Auth header |
|---|---|---|
| OpenAI | https://llm.crabglamp.dev/openai/v1 | Authorization: Bearer cg-pk-… |
| Anthropic | https://llm.crabglamp.dev/anthropic/v1 | x-api-key: cg-pk-… (or Authorization: Bearer cg-pk-…) |
| ElevenLabs | https://llm.crabglamp.dev/elevenlabs | xi-api-key: cg-pk-… |
The proxy accepts any of Authorization: Bearer, x-api-key, or xi-api-key — pick whichever is the SDK default for the upstream provider you are calling.
Supported models and prices
Prices are what you pay, per 1M tokens, reviewed monthly. The proxy passes through any model the provider offers; a model not listed below is billed at the provider's published rate.
Chat models — OpenAI:
| Model | Input ($/1M tokens) | Output ($/1M tokens) |
|---|---|---|
| gpt-5.4 | 3.75 | 22.50 |
| gpt-5.4-mini | 1.125 | 6.75 |
| gpt-5.4-nano | 0.30 | 1.875 |
| gpt-5.4-pro | 45.00 | 270.00 |
| gpt-5.3-codex | 2.625 | 21.00 |
| gpt-4.1 | 3.00 | 12.00 |
| gpt-4.1-mini | 0.60 | 2.40 |
| gpt-4.1-nano | 0.15 | 0.60 |
| gpt-4o | 3.75 | 15.00 |
| gpt-4o-mini | 0.225 | 0.90 |
| o3 | 3.00 | 12.00 |
| o4-mini | 1.65 | 6.60 |
| o3-mini | 1.65 | 6.60 |
Chat models — Anthropic:
| Model | Input ($/1M tokens) | Output ($/1M tokens) |
|---|---|---|
| claude-opus-4-6 | 7.50 | 37.50 |
| claude-sonnet-4-6 | 4.50 | 22.50 |
| claude-opus-4-5 | 7.50 | 37.50 |
| claude-sonnet-4-5 | 4.50 | 22.50 |
| claude-haiku-4-5 | 1.50 | 7.50 |
| claude-opus-4-1 | 22.50 | 112.50 |
| claude-opus-4 | 22.50 | 112.50 |
| claude-sonnet-4 | 4.50 | 22.50 |
Voice — ElevenLabs (per 1k characters):
| Model | Price ($/1k chars) |
|---|---|
| eleven_v3 | 0.18 |
| eleven_multilingual_v2 | 0.18 |
| eleven_flash_v2_5 | 0.09 |
| eleven_flash_v2 | 0.09 |
Spend-cap math
The cap is the maximum a key may spend across LLM and voice combined. Once your running spend reaches the cap, the next request is rejected with HTTP 429 (Monthly spend limit reached). Enforcement is best-effort: spend is checked against a total that refreshes every couple of minutes, so a request or two may slip through slightly over the cap.
CLI
On-VM CLI for virtual keys:
| Command | Purpose |
|---|---|
crabglamp keys create | Create this Agent's virtual key and point OpenClaw at the proxy |
crabglamp keys status | Show the key's status and this month's spend |
crabglamp keys configure | Switch the provider/model the key routes to (OpenClaw) |
crabglamp keys regenerate | Rotate the token; the old token stops working immediately |
crabglamp keys revoke | Revoke the key (regenerate to restore) |
Limits
| Limit | Value |
|---|---|
| Minimum spend cap | $20 |
| Keys per Agent | 1 |
| Usage reporting cadence | within ~2 minutes |
| Token prefix | cg-pk- |