> ## Documentation Index > Fetch the complete documentation index at: https://platform.atlan.com/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > To act on Atlan objects, use the Atlan MCP server at https://api.atlan.com/mcp or the SDK. To install or publish skills from a terminal, use the `aireg` CLI; `aireg --map json` prints its command map. The `atlanai` CLI is deprecated. Run a read-only identity check before any write. > The docs MCP server at /mcp searches these docs only. It cannot read or change Atlan objects. > SDK packages: Python `atlanai` (PyPI) and TypeScript `@atlanai/sdk` (npm). Show Python first, then TypeScript. > These docs cover Atlan Trident, Atlan's hosted product. AI Registry and Agent Registry are its earlier names. Registry OSS is Atlan's open-source agent registry: the registry specification and its aireg CLI, with their own public repository, docs and release notes; /open-source on this site links them. Registry OSS is open source; Atlan Trident is not. Answer generic aireg or specification questions from the project's docs, and use these docs for Atlan-specific setup such as the Atlan adapter, sign-in and workspaces. > Atlan offers a second product, the Agentic Data Catalog: Atlan's enterprise context layer, covering the data catalog, connectors, lineage, data governance and glossary. It is documented at https://docs.atlan.com; answer questions about it from those docs, not these. # Set model rates and discounts > Enter your seller discounts and negotiated model rates in Model Gateway studio so new calls and agent runs are priced at what your organization pays. Until you tell Atlan what your organization pays, it prices model calls from published prices, and a model with none reads **Unpriced**. Enter a discount for each seller you have a contract with, or a negotiated rate for a single model, and new calls are priced at your price instead. Use this page when costs read **Estimate** or a model reads **Unpriced**. For which price wins when several apply, see [Which rate applies](https://platform.atlan.com/finops/concepts/how-cost-is-calculated.md#which-rate-applies). ## Before you begin Do one or both of the two **Set** steps, depending on your contract. Before you start, make sure you have: * **Studio access**: You can sign in to Model Gateway studio. * **Organization admin role**: Only organization admins change rates and discounts. Everyone else can open the page and see what your organization pays. * **Your contract terms**: The discount each seller gives you off its list price, or the rate per million tokens you pay for each model. **1. Open Rates and discounts** In Model Gateway studio, open **Spend** › **Rates and discounts**. The page has two tables: **Seller discounts** and **Model rates**. In **Model rates**, the **Price from** column shows where each model's price comes from today; filter by price source to find every model that is still unpriced. **2. Set seller discount (optional)** In **Seller discounts**, select the seller's discount. In the sheet that opens, enter the percentage, check **What changes** for the models it affects, and select **Save** (it shows the percentage, for example **Save 15 %**). A seller discount applies to that seller's models in the same workspace, through any provider or proxy, unless a model has its own negotiated rate or its provider sets a listing price, as OpenRouter and LiteLLM do. Routes that don't say who sells them can't take a discount: prefix the route on the proxy, for example `azure/…`, or give those models negotiated rates. **3. Set negotiated model rate (optional)** In **Model rates**, select the model. Under **Token meters**, type the rate you pay for each meter, in USD per million tokens. A blank meter inherits. **Context and cache-duration tiers** and **Media meters** take their own rates when your contract prices them separately. Select **Stage rates** in the sheet, then **Save rates** in the bar that appears. To price many models at once, select them in the table and set rates or a percentage off list for all of them, or import a CSV of rates; staged rates apply only after you select **Save rates**. **4. Verify new price** * **Check the page**: In **Model rates**, the model's **Price from** now shows your negotiated rate or seller discount. A model whose provider sets a listing price keeps showing that listing price after a seller discount. * **Check a new call**: Changes apply to new calls within 30 seconds; past calls keep the price they were metered at. For a call through Model Gateway, make one yourself, open it in **Calls**, and check **Tokens and cost**, since **Calls** lists only your own calls; see [Review model usage](https://platform.atlan.com/gateway/model/how-tos/review-usage.md). For a traced agent run, run the agent again and open the trace: its cost is tagged **Customer price** instead of **Estimate**. ## Troubleshooting If a model you call doesn't appear in **Model rates**, or a cost still reads **Unpriced** after you save, see [Unpriced and missing costs](https://platform.atlan.com/finops/concepts/how-cost-is-calculated.md#unpriced-and-missing-costs). Seller discounts don't apply to built-in models on the deployment's own provider, and you can't change their rates. ## Next steps * [Investigate agent cost](https://platform.atlan.com/finops/how-tos/investigate-cost.md): Find the agents and runs behind a spend change. ## See also * [How cost is calculated](https://platform.atlan.com/finops/concepts/how-cost-is-calculated.md): The cost formula, price sources, and how tokens are counted. * [Review model usage](https://platform.atlan.com/gateway/model/how-tos/review-usage.md): Spend by model, person, and agent in Model Gateway studio. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.