Browserbase Model Gateway
Routes a Stagehand agent's model calls to OpenAI, Anthropic and Gemini through the Browserbase API key, handling billing, retries and rate limits, with tokens charged at the providers' own prices.
Find alternatives to Browserbase Model Gateway
What it does
- Model inference
Features 3
- One key for every modelAccess Anthropic, OpenAI, Gemini and more through the existing Browserbase API key.
- Config-level model switchingSwitch models in the Stagehand config without touching credentials or accounts.
- Platform-level rate limitsRate limits are handled by the platform rather than by each team.
Sources
- Pricing
-
Company disclosure · 10 Oct 2026
"We charge market price for tokens, which is the same as going direct." Plan cards: "Model Gateway | Market Price: Pay as you go" on paid plans; the Free card lists "$5 in tokens" without saying whether it recurs, so no free model is recorded.
- Description
-
Company disclosure · 10 Oct 2026
Launch post (2026-04-06): Browserbase routes model requests to the upstream provider and handles billing, retries and rate limits; "Model Gateway currently supports OpenAI, Anthropic, and Gemini."
- What it does
-
Official documentation · 10 Oct 2026
"Access Anthropic, OpenAI, Gemini, and more models through your existing Browserbase API key."
- Sold within
-
Official documentation · 10 Oct 2026
"If you have a Browserbase API key, you're already in."
- Name
-
Official documentation · 10 Oct 2026
"Model Gateway sits between your Stagehand agents and every model they need."
- URL
-
Official documentation · 10 Oct 2026
Product page "Model Gateway | Browserbase", in the site menu as Models.
- Deployment
-
Official documentation · 10 Oct 2026
Hosted gateway.
- Interfaces
-
Official documentation · 10 Oct 2026
Configured in Stagehand ("Swap models in a single line"); "No new SDK. No new authentication flow."
- Intended for
-
Official documentation · 10 Oct 2026
Developers building Stagehand agents.
Also from Browserbase
-
Browserbase Infrastructure service
Cloud browsers, search, fetch and a managed runtime that AI agents use to browse, fill forms and act on websites, with session replays for debugging.
-
Browserbase Agents AI agent
Managed browser agent created from a natural-language prompt and run with one API call; it navigates sites, fills forms and returns structured JSON, with browsers, search, verified identity and proxies included.
-
Browserbase Search API API service
Real-time web search API for agents that returns ranked, token-optimised results (URL, title and favicon per page) so an agent can pick which sites to fetch or browse next.
-
Browserbase Fetch API API service
API that reads a URL live at request time and returns it as clean markdown, HTML or JSON for an agent's context, with ads and navigation stripped and an Extract option for structured fields.
-
Browserbase Runtime Infrastructure service
Sandboxed, isolated hosting for agent code deployed to Browserbase (Browserbase Functions), so an agent runs next to the browser session it controls instead of on a separate server.
-
Browserbase Identity Infrastructure service
Gives each Browserbase session a cryptographically verified agent credential through Web Bot Auth, so partner sites and anti-bot providers let the agent in, with CAPTCHA solving and reusable logged-in contexts.
-
Browserbase Observability Infrastructure service
Debugging view over every Browserbase agent session: live view, action-by-action replay, structured logs of network requests, console output and errors, and traces of each agent decision.
-
Director Application
Director turns plain-English instructions into browser agents that can run on a schedule.
Open Browserbase in the directory
Something wrong here? Send a correction — quote this product id: browserbase-model-gateway.