Lobstack18 models, one endpoint, one bill.
One key and one OpenAI-compatible URL. Every model comes from a US company: Anthropic, OpenAI, Google and xAI.
Or don’t pick one
Nex 1 is Lobstack’s router. Send auto and it picks a model from the catalogue below, never above the tier your plan reaches.
curl https://www.lobstack.ai/api/gateway/v1/chat/completions \ -H "Authorization: Bearer $LOBSTACK_API_KEY" \ -d '{ "model": "auto", "messages": [...] }'
It has no rate of its own. A call through Nex costs whatever it routed to, and the receipt names that model and its price.
What the Gateway serves, and what each million tokens costs.
Rates are what you pay on our keys: each provider’s own price, checked on 16 October 2026, plus Lobstack’s 50%. On your own provider key, your provider bills the tokens.
| Model | Company | Tier | Context | $ / M in | $ / M out |
|---|---|---|---|---|---|
GPT-6 Lunagpt-6-luna | OpenAI | Nano | 1.05M | $0.150 | $0.750 |
Gemini 3.1 Flash Litegemini-3.1-flash-lite | Nano | 1.05M | $0.375 | $2.25 | |
Gemini 3.5 Flash Litegemini-3.5-flash-lite | Nano | 1.05M | $0.450 | $3.75 | |
Claude Haiku 5.5claude-haiku-5-5 | Anthropic | Small | 1M | $0.150 | $0.750 |
Grok Build 0.1grok-build-0.1 | xAI | Small | 256K | $1.50 | $3.00 |
Gemini 3.8 Flashgemini-3.8-flash | Standard | 1.05M | $1.13 | $5.63 | |
Grok 4.3grok-4.3 | xAI | Standard | 1M | $1.88 | $3.75 |
Claude Sonnet 5.5claude-sonnet-5-5 | Anthropic | Standard | 1M | $3.00 | $15.00 |
Gemini 3.1 Progemini-3.1-pro | Premium | 1.05M | $3.00 | $18.00 | |
GPT-6 Solgpt-6-sol | OpenAI | Premium | 1.05M | $3.00 | $15.00 |
GPT-6.1 Solgpt-6.1-sol | OpenAI | Premium | 1.05M | $3.00 | $15.00 |
Grok 4.7grok-4.7 | xAI | Premium | 500K | $3.00 | $9.00 |
Grok 4.6grok-4.6 | xAI | Premium | 500K | $3.00 | $9.00 |
Grok 4.5grok-4.5 | xAI | Premium | 500K | $3.00 | $9.00 |
Gemini 4 Argongemini-4-argon | Flagship | 1.05M | $3.00 | $15.00 | |
Claude Opus 5.5claude-opus-5-5 | Anthropic | Flagship | 1M | $6.00 | $30.00 |
Claude Fable 5.1claude-fable-5-1 | Anthropic | Flagship | 1M | $15.00 | $75.00 |
GPT-6 Astragpt-6-astra | OpenAI | Flagship | 1.05M | $15.00 | $75.00 |
Older models keep their model strings. Two now forward to a newer model, shown with an arrow.
How a rate becomes a line on your bill.
01
Per call
Tokens in times the input rate, tokens out times the output rate, on the model the response names.
02
Against your plan
That cost draws down the model spend included in your plan, and you top up when it runs out.
03
Your own keys
On paid plans, calls on your own provider keys skip these rates and your allowance; your provider bills the tokens.
Questions, answered.
Yes. Each rate is the provider’s own list price plus 50%, rounded, so the number in the table is what a call on our keys costs. There is no separate fee.
It sends the call to Nex 1, Lobstack’s router. Nex scores the request and serves it from the cheapest tier that can handle it, never above your plan’s tier.
Yes, on a paid plan. Calls on your own key skip these rates and your plan’s allowance, and your provider bills the tokens directly.
Lobstack is an American company, and every model it serves comes from a US company: Anthropic, OpenAI, Google and xAI. Each is named on the receipt.
A price band, from Nano to Flagship, that groups models of similar cost. Your plan sets the highest tier Nex will route to.
Pick a model, or send auto.
Free to start. A verified account gets $5 of Gateway credit once for 30 days, whether it starts with the API or the app. No card.