Grandet
Grandet compares LLM API prices by model across official, inference-provider and relay sites, for developers and AI assistants hunting for the cheapest API.
No visits from OpenBeat in the last 30 days yet. Every click on “Visit” is counted here.
Pricing and API checked by an OpenBeat editor on Nov 12, 2026 · facts from the maker’s own site ·
Tool insights
Grandet lists what different sellers charge for the same LLM, so you can see official, inference-provider and relay prices side by side. It also offers a price API and an MCP server, so assistants and scripts can look up prices without a manual search.
The same model is often sold by its maker, by inference providers and by relay sites at different prices and in different units. Grandet puts these into one comparison so you don't have to check each site yourself.
Developers, and people who buy LLM API access and want to find the cheapest source for a given model. It is also built for AI assistants that need price data. If you need verified cash prices for relay sites, it may not be enough, because the site says it does not check that.
What it does
Grandet is a price comparison site for LLM APIs. You pick a model and see what it costs across three kinds of sellers: the official maker, inference providers and relay sites. The maker of Grandet is not named on the site.
The model list in the facts includes claude-sonnet-5-5, gpt-6-astra, gemini-4-argon, glm-5.3, mistral-large-4, DeepSeek-V4.1-Flash, grok-4.7, qwen3.7-plus-2026-05-26 and kimi-k3.1-pro.
Key features
- Official, inference-provider and relay prices are listed separately.
- Relay prices are converted to a price per million tokens.
- Each price shows a group name and the date it was observed.
- A price API endpoint returns data per model.
- A free credits list covers official and relay offers.
- An anonymous MCP server works with no key.
- A Grandet Skill is available for AI assistants.
Pricing
The pricing model, free plan and paid plans are not stated on the maker's site, as of November 2026. No starting price is given. The site's leaderboard page is the place to check the current state.
API and integrations
Grandet has an API, including a price endpoint per model. Its MCP server needs no key. The listed integrations are Codex, Claude Code, Cursor and MCP. No SDKs are stated. Platforms are the web and MCP.
Good to know
- Relay prices are site credit, not cash. Grandet compares them separately from other prices.
- It does not check whether relay prices can be bought with real money.
- Many relay sites could not be reached in the latest update.
- Observations cover 2026-10-03 to 2026-11-07.
- The latest data update on 2026-11-07 (v1.8.0) collected 269,678 API prices.
- Whether the project is open source, and under what license, is not stated.
- Documentation is published as an llms.txt file, and a changelog is on the site.
FAQ
What does Grandet do?
Grandet compares LLM API prices by model across official sites, inference providers and relay sites. It lists each group separately and shows the date each price was observed.
Is Grandet free?
As of November 2026, the maker's site does not state a pricing model, a free plan or any paid plans. It does offer a list of free credits from official and relay providers, which is a separate thing from its own pricing.
Does Grandet have an API or MCP server?
Yes. Grandet has a price API endpoint per model and an anonymous MCP server that needs no key. It also lists integrations with Codex, Claude Code and Cursor. No SDKs are stated.
Can I trust relay prices on Grandet?
Treat them with care. Relay prices are site credit, not cash, so Grandet compares them separately and converts them to a price per million tokens. It does not check whether they can be bought with real money, and many relay sites were unreachable in the latest update.
How current is Grandet's price data?
The latest data update is dated 2026-11-07 and collected 269,678 API prices. Observations cover 2026-10-03 to 2026-11-07, and the site shows an observation date for each price.
Releases
No tracked releases yet.
In the news
No coverage yet.