diff --git a/admin/assets/javascripts/discourse/routes/admin-plugins-show-discourse-ai-llms-edit.js b/admin/assets/javascripts/discourse/routes/admin-plugins-show-discourse-ai-llms-edit.js
index 8230211c..f7b5de4c 100644
--- a/admin/assets/javascripts/discourse/routes/admin-plugins-show-discourse-ai-llms-edit.js
+++ b/admin/assets/javascripts/discourse/routes/admin-plugins-show-discourse-ai-llms-edit.js
@@ -2,8 +2,17 @@ import DiscourseRoute from "discourse/routes/discourse";
 
 export default class AdminPluginsShowDiscourseAiLlmsEdit extends DiscourseRoute {
   async model(params) {
-    const allLlms = this.modelFor("adminPlugins.show.discourse-ai-llms");
     const id = parseInt(params.id, 10);
+
+    if (id < 0) {
+      // You shouldn't be able to access the edit page
+      // if the model is seeded
+      return this.router.transitionTo(
+        "adminPlugins.show.discourse-ai-llms.index"
+      );
+    }
+
+    const allLlms = this.modelFor("adminPlugins.show.discourse-ai-llms");
     const record = allLlms.findBy("id", id);
     record.provider_params = record.provider_params || {};
     return record;
diff --git a/app/controllers/discourse_ai/admin/ai_llms_controller.rb b/app/controllers/discourse_ai/admin/ai_llms_controller.rb
index 68ddcf94..84109220 100644
--- a/app/controllers/discourse_ai/admin/ai_llms_controller.rb
+++ b/app/controllers/discourse_ai/admin/ai_llms_controller.rb
@@ -161,6 +161,9 @@ module DiscourseAi
             :api_key,
             :enabled_chat_bot,
             :vision_enabled,
+            :input_cost,
+            :cached_input_cost,
+            :output_cost,
           )
 
         provider = updating ? updating.provider : permitted[:provider]
diff --git a/app/models/llm_model.rb b/app/models/llm_model.rb
index 111ac70a..05c7be4b 100644
--- a/app/models/llm_model.rb
+++ b/app/models/llm_model.rb
@@ -13,7 +13,14 @@ class LlmModel < ActiveRecord::Base
   validates :url, presence: true, unless: -> { provider == BEDROCK_PROVIDER_NAME }
   validates_presence_of :name, :api_key
   validates :max_prompt_tokens, numericality: { greater_than: 0 }
-  validates :max_output_tokens, numericality: { greater_than_or_equal_to: 0 }, allow_nil: true
+  validates :input_cost,
+            :cached_input_cost,
+            :output_cost,
+            :max_output_tokens,
+            numericality: {
+              greater_than_or_equal_to: 0,
+            },
+            allow_nil: true
   validate :required_provider_params
   scope :in_use,
         -> do
@@ -184,5 +191,8 @@ end
 #  enabled_chat_bot  :boolean          default(FALSE), not null
 #  provider_params   :jsonb
 #  vision_enabled    :boolean          default(FALSE), not null
+#  input_cost        :float
+#  cached_input_cost :float
+#  output_cost       :float
 #  max_output_tokens :integer
 #
diff --git a/app/serializers/ai_usage_serializer.rb b/app/serializers/ai_usage_serializer.rb
index 38969b05..0a990547 100644
--- a/app/serializers/ai_usage_serializer.rb
+++ b/app/serializers/ai_usage_serializer.rb
@@ -22,6 +22,9 @@ class AiUsageSerializer < ApplicationSerializer
         total_cached_tokens
         total_request_tokens
         total_response_tokens
+        input_spending
+        output_spending
+        cached_input_spending
       ],
     )
   end
@@ -35,6 +38,9 @@ class AiUsageSerializer < ApplicationSerializer
         total_cached_tokens
         total_request_tokens
         total_response_tokens
+        input_spending
+        output_spending
+        cached_input_spending
       ],
     )
   end
@@ -49,6 +55,9 @@ class AiUsageSerializer < ApplicationSerializer
         total_cached_tokens: user.total_cached_tokens,
         total_request_tokens: user.total_request_tokens,
         total_response_tokens: user.total_response_tokens,
+        input_spending: user.input_spending,
+        output_spending: user.output_spending,
+        cached_input_spending: user.cached_input_spending,
       }
     end
   end
@@ -60,6 +69,7 @@ class AiUsageSerializer < ApplicationSerializer
       total_request_tokens: object.total_request_tokens,
       total_response_tokens: object.total_response_tokens,
       total_requests: object.total_requests,
+      total_spending: object.total_spending,
       date_range: {
         start: object.start_date,
         end: object.end_date,
diff --git a/app/serializers/llm_model_serializer.rb b/app/serializers/llm_model_serializer.rb
index 33e62d8b..736757d4 100644
--- a/app/serializers/llm_model_serializer.rb
+++ b/app/serializers/llm_model_serializer.rb
@@ -18,6 +18,9 @@ class LlmModelSerializer < ApplicationSerializer
              :enabled_chat_bot,
              :provider_params,
              :vision_enabled,
+             :input_cost,
+             :output_cost,
+             :cached_input_cost,
              :used_by
 
   has_one :user, serializer: BasicUserSerializer, embed: :object
diff --git a/assets/javascripts/discourse/admin/models/ai-llm.js b/assets/javascripts/discourse/admin/models/ai-llm.js
index c434ba25..cafbf3a0 100644
--- a/assets/javascripts/discourse/admin/models/ai-llm.js
+++ b/assets/javascripts/discourse/admin/models/ai-llm.js
@@ -15,7 +15,10 @@ export default class AiLlm extends RestModel {
       "api_key",
       "enabled_chat_bot",
       "provider_params",
-      "vision_enabled"
+      "vision_enabled",
+      "input_cost",
+      "cached_input_cost",
+      "output_cost"
     );
   }
 
diff --git a/assets/javascripts/discourse/components/ai-llm-editor-form.gjs b/assets/javascripts/discourse/components/ai-llm-editor-form.gjs
index 50ab2ff7..38cf1b50 100644
--- a/assets/javascripts/discourse/components/ai-llm-editor-form.gjs
+++ b/assets/javascripts/discourse/components/ai-llm-editor-form.gjs
@@ -47,6 +47,9 @@ export default class AiLlmEditorForm extends Component {
         name: modelInfo.name,
         provider: info.provider,
         provider_params: this.computeProviderParams(info.provider),
+        input_cost: modelInfo.input_cost,
+        output_cost: modelInfo.output_cost,
+        cached_input_cost: modelInfo.cached_input_cost,
       };
     }
 
@@ -63,6 +66,9 @@ export default class AiLlmEditorForm extends Component {
       provider: model.provider,
       enabled_chat_bot: model.enabled_chat_bot,
       vision_enabled: model.vision_enabled,
+      input_cost: model.input_cost,
+      output_cost: model.output_cost,
+      cached_input_cost: model.cached_input_cost,
       provider_params: this.computeProviderParams(
         model.provider,
         model.provider_params
@@ -118,10 +124,6 @@ export default class AiLlmEditorForm extends Component {
     return localized.join(", ");
   }
 
-  get seeded() {
-    return this.args.model.id < 0;
-  }
-
   get inUseWarning() {
     return i18n("discourse_ai.llms.in_use_warning", {
       settings: this.modulesUsingModel,
@@ -271,15 +273,9 @@ export default class AiLlmEditorForm extends Component {
     <Form
       @onSubmit={{this.save}}
       @data={{this.formData}}
-      class="ai-llm-editor {{if this.seeded 'seeded'}}"
+      class="ai-llm-editor"
       as |form data|
     >
-      {{#if this.seeded}}
-        <form.Alert @icon="circle-info">
-          {{i18n "discourse_ai.llms.seeded_warning"}}
-        </form.Alert>
-      {{/if}}
-
       {{#if this.modulesUsingModel}}
         <form.Alert @icon="circle-info">
           {{this.inUseWarning}}
@@ -290,7 +286,6 @@ export default class AiLlmEditorForm extends Component {
         @name="display_name"
         @title={{i18n "discourse_ai.llms.display_name"}}
         @validation="required|length:1,100"
-        @disabled={{this.seeded}}
         @format="large"
         @tooltip={{i18n "discourse_ai.llms.hints.display_name"}}
         as |field|
@@ -303,7 +298,6 @@ export default class AiLlmEditorForm extends Component {
         @title={{i18n "discourse_ai.llms.name"}}
         @tooltip={{i18n "discourse_ai.llms.hints.name"}}
         @validation="required"
-        @disabled={{this.seeded}}
         @format="large"
         as |field|
       >
@@ -313,7 +307,6 @@ export default class AiLlmEditorForm extends Component {
       <form.Field
         @name="provider"
         @title={{i18n "discourse_ai.llms.provider"}}
-        @disabled={{this.seeded}}
         @format="large"
         @validation="required"
         @onSet={{this.setProvider}}
@@ -328,262 +321,289 @@ export default class AiLlmEditorForm extends Component {
         </field.Select>
       </form.Field>
 
-      {{#unless this.seeded}}
-        {{#if (this.canEditURL data.provider)}}
-          <form.Field
-            @name="url"
-            @title={{i18n "discourse_ai.llms.url"}}
-            @validation="required"
-            @format="large"
-            as |field|
-          >
-            <field.Input />
-          </form.Field>
-        {{/if}}
-
+      {{#if (this.canEditURL data.provider)}}
         <form.Field
-          @name="api_key"
-          @title={{i18n "discourse_ai.llms.api_key"}}
+          @name="url"
+          @title={{i18n "discourse_ai.llms.url"}}
           @validation="required"
           @format="large"
           as |field|
         >
-          <field.Password />
+          <field.Input />
         </form.Field>
+      {{/if}}
 
-        <form.Object @name="provider_params" as |object providerParamsData|>
-          {{#each (this.providerParamsKeys providerParamsData) as |name|}}
-            {{#let
-              (get (this.metaProviderParams data.provider) name)
-              as |params|
-            }}
-              <object.Field
-                @name={{name}}
-                @title={{i18n
-                  (concat "discourse_ai.llms.provider_fields." name)
-                }}
-                @format="large"
-                as |field|
-              >
-                {{#if (eq params.type "enum")}}
-                  <field.Select @includeNone={{false}} as |select|>
-                    {{#each params.values as |option|}}
-                      <select.Option
-                        @value={{option.id}}
-                      >{{option.name}}</select.Option>
-                    {{/each}}
-                  </field.Select>
-                {{else if (eq params.type "checkbox")}}
-                  <field.Checkbox />
-                {{else}}
-                  <field.Input @type={{params.type}} />
-                {{/if}}
-              </object.Field>
-            {{/let}}
-          {{/each}}
-        </form.Object>
+      <form.Field
+        @name="api_key"
+        @title={{i18n "discourse_ai.llms.api_key"}}
+        @validation="required"
+        @format="large"
+        as |field|
+      >
+        <field.Password />
+      </form.Field>
 
-        <form.Field
-          @name="tokenizer"
-          @title={{i18n "discourse_ai.llms.tokenizer"}}
-          @disabled={{this.seeded}}
-          @format="large"
-          @validation="required"
-          as |field|
-        >
-          <field.Select as |select|>
-            {{#each this.tokenizers as |tokenizer|}}
-              <select.Option
-                @value={{tokenizer.id}}
-              >{{tokenizer.name}}</select.Option>
-            {{/each}}
-          </field.Select>
-        </form.Field>
-
-        <form.Field
-          @name="max_prompt_tokens"
-          @title={{i18n "discourse_ai.llms.max_prompt_tokens"}}
-          @tooltip={{i18n "discourse_ai.llms.hints.max_prompt_tokens"}}
-          @validation="required"
-          @format="large"
-          as |field|
-        >
-          <field.Input @type="number" step="any" min="0" lang="en" />
-        </form.Field>
-
-        <form.Field
-          @name="max_output_tokens"
-          @title={{i18n "discourse_ai.llms.max_output_tokens"}}
-          @tooltip={{i18n "discourse_ai.llms.hints.max_output_tokens"}}
-          @format="large"
-          as |field|
-        >
-          <field.Input @type="number" step="any" min="0" lang="en" />
-        </form.Field>
-
-        <form.Field
-          @name="vision_enabled"
-          @title={{i18n "discourse_ai.llms.vision_enabled"}}
-          @tooltip={{i18n "discourse_ai.llms.hints.vision_enabled"}}
-          @format="large"
-          as |field|
-        >
-          <field.Checkbox />
-        </form.Field>
-
-        <form.Field
-          @name="enabled_chat_bot"
-          @title={{i18n "discourse_ai.llms.enabled_chat_bot"}}
-          @tooltip={{i18n "discourse_ai.llms.hints.enabled_chat_bot"}}
-          @format="large"
-          as |field|
-        >
-          <field.Checkbox />
-        </form.Field>
-
-        {{#if @model.user}}
-          <form.Container @title={{i18n "discourse_ai.llms.ai_bot_user"}}>
-            <a
-              class="avatar"
-              href={{@model.user.path}}
-              data-user-card={{@model.user.username}}
+      <form.Object @name="provider_params" as |object providerParamsData|>
+        {{#each (this.providerParamsKeys providerParamsData) as |name|}}
+          {{#let
+            (get (this.metaProviderParams data.provider) name)
+            as |params|
+          }}
+            <object.Field
+              @name={{name}}
+              @title={{i18n (concat "discourse_ai.llms.provider_fields." name)}}
+              @format="large"
+              as |field|
             >
-              {{Avatar @model.user.avatar_template "small"}}
-            </a>
-            <LinkTo @route="adminUser" @model={{this.adminUser}}>
-              {{@model.user.username}}
-            </LinkTo>
-          </form.Container>
-        {{/if}}
+              {{#if (eq params.type "enum")}}
+                <field.Select @includeNone={{false}} as |select|>
+                  {{#each params.values as |option|}}
+                    <select.Option
+                      @value={{option.id}}
+                    >{{option.name}}</select.Option>
+                  {{/each}}
+                </field.Select>
+              {{else if (eq params.type "checkbox")}}
+                <field.Checkbox />
+              {{else}}
+                <field.Input @type={{params.type}} />
+              {{/if}}
+            </object.Field>
+          {{/let}}
+        {{/each}}
+      </form.Object>
 
-        {{#if (gt data.llm_quotas.length 0)}}
-          <form.Container @title={{i18n "discourse_ai.llms.quotas.title"}}>
-            <table class="ai-llm-quotas__table">
-              <thead class="ai-llm-quotas__table-head">
-                <tr class="ai-llm-quotas__header-row">
-                  <th class="ai-llm-quotas__header">{{i18n
-                      "discourse_ai.llms.quotas.group"
-                    }}</th>
-                  <th class="ai-llm-quotas__header">{{i18n
-                      "discourse_ai.llms.quotas.max_tokens"
-                    }}</th>
-                  <th class="ai-llm-quotas__header">{{i18n
-                      "discourse_ai.llms.quotas.max_usages"
-                    }}</th>
-                  <th class="ai-llm-quotas__header">{{i18n
-                      "discourse_ai.llms.quotas.duration"
-                    }}</th>
-                  <th
-                    class="ai-llm-quotas__header ai-llm-quotas__header--actions"
-                  ></th>
-                  <th></th>
-                </tr>
-              </thead>
-              <tbody class="ai-llm-quotas__table-body">
-                <form.Collection
-                  @name="llm_quotas"
-                  @tagName="tr"
-                  class="ai-llm-quotas__row"
-                  as |collection index collectionData|
-                >
-                  <td
-                    class="ai-llm-quotas__cell"
-                  >{{collectionData.group_name}}</td>
-                  <td class="ai-llm-quotas__cell">
-                    <collection.Field
-                      @name="max_tokens"
-                      @title="max_tokens"
-                      @showTitle={{false}}
-                      as |field|
-                    >
-                      <field.Input
-                        @type="number"
-                        class="ai-llm-quotas__input"
-                        min="1"
-                      />
-                    </collection.Field>
-                  </td>
-                  <td class="ai-llm-quotas__cell">
-                    <collection.Field
-                      @name="max_usages"
-                      @title="max_usages"
-                      @showTitle={{false}}
-                      as |field|
-                    >
-                      <field.Input
-                        @type="number"
-                        class="ai-llm-quotas__input"
-                        min="1"
-                      />
-                    </collection.Field>
-                  </td>
-                  <td class="ai-llm-quotas__cell">
-                    <collection.Field
-                      @name="duration_seconds"
-                      @title="duration_seconds"
-                      @showTitle={{false}}
-                      as |field|
-                    >
-                      <field.Custom>
-                        <DurationSelector
-                          @value={{collectionData.duration_seconds}}
-                          @onChange={{field.set}}
-                        />
-                      </field.Custom>
-                    </collection.Field>
-                  </td>
-                  <td>
-                    <form.Button
-                      @icon="trash-can"
-                      @action={{fn collection.remove index}}
-                      class="btn-danger ai-llm-quotas__delete-btn"
+      <form.Field
+        @name="tokenizer"
+        @title={{i18n "discourse_ai.llms.tokenizer"}}
+        @format="large"
+        @validation="required"
+        as |field|
+      >
+        <field.Select as |select|>
+          {{#each this.tokenizers as |tokenizer|}}
+            <select.Option
+              @value={{tokenizer.id}}
+            >{{tokenizer.name}}</select.Option>
+          {{/each}}
+        </field.Select>
+      </form.Field>
+
+      <form.Field
+        @name="max_prompt_tokens"
+        @title={{i18n "discourse_ai.llms.max_prompt_tokens"}}
+        @tooltip={{i18n "discourse_ai.llms.hints.max_prompt_tokens"}}
+        @validation="required"
+        @format="large"
+        as |field|
+      >
+        <field.Input @type="number" step="any" min="0" lang="en" />
+      </form.Field>
+
+      <form.InputGroup as |inputGroup|>
+        <inputGroup.Field
+          @name="input_cost"
+          @title={{i18n "discourse_ai.llms.cost_input"}}
+          @tooltip={{i18n "discourse_ai.llms.hints.cost_input"}}
+          @helpText={{i18n "discourse_ai.llms.hints.cost_measure"}}
+          as |field|
+        >
+          <field.Input @type="number" step="any" min="0" lang="en" />
+        </inputGroup.Field>
+
+        <inputGroup.Field
+          @name="cached_input_cost"
+          @title={{i18n "discourse_ai.llms.cost_cached_input"}}
+          @tooltip={{i18n "discourse_ai.llms.hints.cost_cached_input"}}
+          @helpText={{i18n "discourse_ai.llms.hints.cost_measure"}}
+          as |field|
+        >
+          <field.Input @type="number" step="any" min="0" lang="en" />
+        </inputGroup.Field>
+
+        <inputGroup.Field
+          @name="output_cost"
+          @title={{i18n "discourse_ai.llms.cost_output"}}
+          @tooltip={{i18n "discourse_ai.llms.hints.cost_output"}}
+          @helpText={{i18n "discourse_ai.llms.hints.cost_measure"}}
+          as |field|
+        >
+          <field.Input @type="number" step="any" min="0" lang="en" />
+        </inputGroup.Field>
+      </form.InputGroup>
+
+      <form.Field
+        @name="max_output_tokens"
+        @title={{i18n "discourse_ai.llms.max_output_tokens"}}
+        @tooltip={{i18n "discourse_ai.llms.hints.max_output_tokens"}}
+        @format="large"
+        as |field|
+      >
+        <field.Input @type="number" step="any" min="0" lang="en" />
+      </form.Field>
+
+      <form.Field
+        @name="vision_enabled"
+        @title={{i18n "discourse_ai.llms.vision_enabled"}}
+        @tooltip={{i18n "discourse_ai.llms.hints.vision_enabled"}}
+        @format="large"
+        as |field|
+      >
+        <field.Checkbox />
+      </form.Field>
+
+      <form.Field
+        @name="enabled_chat_bot"
+        @title={{i18n "discourse_ai.llms.enabled_chat_bot"}}
+        @tooltip={{i18n "discourse_ai.llms.hints.enabled_chat_bot"}}
+        @format="large"
+        as |field|
+      >
+        <field.Checkbox />
+      </form.Field>
+
+      {{#if @model.user}}
+        <form.Container @title={{i18n "discourse_ai.llms.ai_bot_user"}}>
+          <a
+            class="avatar"
+            href={{@model.user.path}}
+            data-user-card={{@model.user.username}}
+          >
+            {{Avatar @model.user.avatar_template "small"}}
+          </a>
+          <LinkTo @route="adminUser" @model={{this.adminUser}}>
+            {{@model.user.username}}
+          </LinkTo>
+        </form.Container>
+      {{/if}}
+
+      {{#if (gt data.llm_quotas.length 0)}}
+        <form.Container @title={{i18n "discourse_ai.llms.quotas.title"}}>
+          <table class="ai-llm-quotas__table">
+            <thead class="ai-llm-quotas__table-head">
+              <tr class="ai-llm-quotas__header-row">
+                <th class="ai-llm-quotas__header">{{i18n
+                    "discourse_ai.llms.quotas.group"
+                  }}</th>
+                <th class="ai-llm-quotas__header">{{i18n
+                    "discourse_ai.llms.quotas.max_tokens"
+                  }}</th>
+                <th class="ai-llm-quotas__header">{{i18n
+                    "discourse_ai.llms.quotas.max_usages"
+                  }}</th>
+                <th class="ai-llm-quotas__header">{{i18n
+                    "discourse_ai.llms.quotas.duration"
+                  }}</th>
+                <th
+                  class="ai-llm-quotas__header ai-llm-quotas__header--actions"
+                ></th>
+                <th></th>
+              </tr>
+            </thead>
+            <tbody class="ai-llm-quotas__table-body">
+              <form.Collection
+                @name="llm_quotas"
+                @tagName="tr"
+                class="ai-llm-quotas__row"
+                as |collection index collectionData|
+              >
+                <td
+                  class="ai-llm-quotas__cell"
+                >{{collectionData.group_name}}</td>
+                <td class="ai-llm-quotas__cell">
+                  <collection.Field
+                    @name="max_tokens"
+                    @title="max_tokens"
+                    @showTitle={{false}}
+                    as |field|
+                  >
+                    <field.Input
+                      @type="number"
+                      class="ai-llm-quotas__input"
+                      min="1"
                     />
-                  </td>
-                </form.Collection>
-              </tbody>
-            </table>
-          </form.Container>
+                  </collection.Field>
+                </td>
+                <td class="ai-llm-quotas__cell">
+                  <collection.Field
+                    @name="max_usages"
+                    @title="max_usages"
+                    @showTitle={{false}}
+                    as |field|
+                  >
+                    <field.Input
+                      @type="number"
+                      class="ai-llm-quotas__input"
+                      min="1"
+                    />
+                  </collection.Field>
+                </td>
+                <td class="ai-llm-quotas__cell">
+                  <collection.Field
+                    @name="duration_seconds"
+                    @title="duration_seconds"
+                    @showTitle={{false}}
+                    as |field|
+                  >
+                    <field.Custom>
+                      <DurationSelector
+                        @value={{collectionData.duration_seconds}}
+                        @onChange={{field.set}}
+                      />
+                    </field.Custom>
+                  </collection.Field>
+                </td>
+                <td>
+                  <form.Button
+                    @icon="trash-can"
+                    @action={{fn collection.remove index}}
+                    class="btn-danger ai-llm-quotas__delete-btn"
+                  />
+                </td>
+              </form.Collection>
+            </tbody>
+          </table>
+        </form.Container>
 
+        <form.Button
+          @action={{fn
+            this.openAddQuotaModal
+            (fn form.addItemToCollection "llm_quotas")
+          }}
+          @icon="plus"
+          @label="discourse_ai.llms.quotas.add"
+          class="ai-llm-editor__add-quota-btn"
+        />
+      {{/if}}
+
+      <form.Actions>
+        <form.Button
+          @action={{fn this.test data}}
+          @disabled={{this.testRunning}}
+          @label="discourse_ai.llms.tests.title"
+        />
+
+        <form.Submit />
+
+        {{#if (eq data.llm_quotas.length 0)}}
           <form.Button
             @action={{fn
               this.openAddQuotaModal
               (fn form.addItemToCollection "llm_quotas")
             }}
-            @icon="plus"
             @label="discourse_ai.llms.quotas.add"
             class="ai-llm-editor__add-quota-btn"
           />
         {{/if}}
 
-        <form.Actions>
+        {{#unless @model.isNew}}
           <form.Button
-            @action={{fn this.test data}}
-            @disabled={{this.testRunning}}
-            @label="discourse_ai.llms.tests.title"
+            @action={{this.delete}}
+            @label="discourse_ai.llms.delete"
+            class="btn-danger"
           />
-
-          <form.Submit />
-
-          {{#if (eq data.llm_quotas.length 0)}}
-            <form.Button
-              @action={{fn
-                this.openAddQuotaModal
-                (fn form.addItemToCollection "llm_quotas")
-              }}
-              @label="discourse_ai.llms.quotas.add"
-              class="ai-llm-editor__add-quota-btn"
-            />
-          {{/if}}
-
-          {{#unless @model.isNew}}
-            <form.Button
-              @action={{this.delete}}
-              @label="discourse_ai.llms.delete"
-              class="btn-danger"
-            />
-          {{/unless}}
-        </form.Actions>
-      {{/unless}}
+        {{/unless}}
+      </form.Actions>
 
       {{#if this.displayTestResult}}
         <form.Container @format="full">
diff --git a/assets/javascripts/discourse/components/ai-usage.gjs b/assets/javascripts/discourse/components/ai-usage.gjs
index a972726b..0aaac6b2 100644
--- a/assets/javascripts/discourse/components/ai-usage.gjs
+++ b/assets/javascripts/discourse/components/ai-usage.gjs
@@ -2,6 +2,8 @@ import Component from "@glimmer/component";
 import { tracked } from "@glimmer/tracking";
 import { fn, hash } from "@ember/helper";
 import { action } from "@ember/object";
+import didInsert from "@ember/render-modifiers/modifiers/did-insert";
+import didUpdate from "@ember/render-modifiers/modifiers/did-update";
 import { LinkTo } from "@ember/routing";
 import { service } from "@ember/service";
 import { eq, gt, lt } from "truth-helpers";
@@ -74,6 +76,22 @@ export default class AiUsage extends Component {
     this.onFilterChange();
   }
 
+  @action
+  addCurrencyChar(element) {
+    element.querySelectorAll(".d-stat-tile__label").forEach((label) => {
+      if (
+        label.innerText.trim() === i18n("discourse_ai.usage.total_spending")
+      ) {
+        const valueElement = label
+          .closest(".d-stat-tile")
+          ?.querySelector(".d-stat-tile__value");
+        if (valueElement) {
+          valueElement.innerText = `$${valueElement.innerText}`;
+        }
+      }
+    });
+  }
+
   @bind
   takeUsers(start, end) {
     return this.data.users.slice(start, end);
@@ -153,6 +171,11 @@ export default class AiUsage extends Component {
         value: this.data.summary.total_cached_tokens,
         tooltip: i18n("discourse_ai.usage.stat_tooltips.cached_tokens"),
       },
+      {
+        label: i18n("discourse_ai.usage.total_spending"),
+        value: this.data.summary.total_spending,
+        tooltip: i18n("discourse_ai.usage.stat_tooltips.total_spending"),
+      },
     ];
   }
 
@@ -308,6 +331,11 @@ export default class AiUsage extends Component {
     this.fetchData();
   }
 
+  totalSpending(inputSpending, cachedSpending, outputSpending) {
+    const total = inputSpending + cachedSpending + outputSpending;
+    return `$${total.toFixed(2)}`;
+  }
+
   <template>
     <div class="ai-usage admin-detail">
       <DPageSubheader
@@ -376,9 +404,15 @@ export default class AiUsage extends Component {
             class="ai-usage__summary"
           >
             <:content>
-              <DStatTiles as |tiles|>
+              <DStatTiles
+                {{didInsert this.addCurrencyChar this.metrics}}
+                {{didUpdate this.addCurrencyChar this.metrics}}
+                as |tiles|
+              >
+
                 {{#each this.metrics as |metric|}}
                   <tiles.Tile
+                    class="bar"
                     @label={{metric.label}}
                     @href={{metric.href}}
                     @value={{metric.value}}
@@ -422,6 +456,7 @@ export default class AiUsage extends Component {
                         <th>{{i18n "discourse_ai.usage.feature"}}</th>
                         <th>{{i18n "discourse_ai.usage.usage_count"}}</th>
                         <th>{{i18n "discourse_ai.usage.total_tokens"}}</th>
+                        <th>{{i18n "discourse_ai.usage.total_spending"}}</th>
                       </tr>
                     </thead>
                     <tbody>
@@ -438,6 +473,13 @@ export default class AiUsage extends Component {
                             class="ai-usage__features-cell"
                             title={{feature.total_tokens}}
                           >{{number feature.total_tokens}}</td>
+                          <td>
+                            {{this.totalSpending
+                              feature.input_spending
+                              feature.cached_input_spending
+                              feature.output_spending
+                            }}
+                          </td>
                         </tr>
                       {{/each}}
                     </tbody>
@@ -464,6 +506,8 @@ export default class AiUsage extends Component {
                         <th>{{i18n "discourse_ai.usage.model"}}</th>
                         <th>{{i18n "discourse_ai.usage.usage_count"}}</th>
                         <th>{{i18n "discourse_ai.usage.total_tokens"}}</th>
+                        <th>{{i18n "discourse_ai.usage.total_spending"}}</th>
+
                       </tr>
                     </thead>
                     <tbody>
@@ -478,6 +522,13 @@ export default class AiUsage extends Component {
                             class="ai-usage__models-cell"
                             title={{model.total_tokens}}
                           >{{number model.total_tokens}}</td>
+                          <td>
+                            {{this.totalSpending
+                              model.input_spending
+                              model.cached_input_spending
+                              model.output_spending
+                            }}
+                          </td>
                         </tr>
                       {{/each}}
                     </tbody>
@@ -511,6 +562,7 @@ export default class AiUsage extends Component {
                           }}</th>
                         <th>{{i18n "discourse_ai.usage.usage_count"}}</th>
                         <th>{{i18n "discourse_ai.usage.total_tokens"}}</th>
+                        <th>{{i18n "discourse_ai.usage.total_spending"}}</th>
                       </tr>
                     </thead>
                     <tbody>
@@ -535,6 +587,13 @@ export default class AiUsage extends Component {
                             class="ai-usage__users-cell"
                             title={{user.total_tokens}}
                           >{{number user.total_tokens}}</td>
+                          <td>
+                            {{this.totalSpending
+                              user.input_spending
+                              user.cached_input_spending
+                              user.output_spending
+                            }}
+                          </td>
                         </tr>
                       {{/each}}
                     </tbody>
diff --git a/config/locales/client.en.yml b/config/locales/client.en.yml
index f50dda7f..1811de61 100644
--- a/config/locales/client.en.yml
+++ b/config/locales/client.en.yml
@@ -239,6 +239,7 @@ en:
         net_request_tokens: "Net request tokens"
         cached_tokens: "Cached tokens"
         cached_request_tokens: "Cached request tokens"
+        total_spending: "Estimated cost"
         no_users: "No user usage data found"
         no_models: "No model usage data found"
         no_features: "No feature usage data found"
@@ -249,6 +250,7 @@ en:
           request_tokens: "Tokens used when the LLM tries to understand what you are saying"
           response_tokens: "Tokens used when the LLM responds to your prompt"
           cached_tokens: "Previously processed request tokens that the LLM reuses to optimize performance and cost"
+          total_spending: "Cumulative cost of all tokens used by the LLMs based on specified cost metrics added to LLM configuration settings"
         periods:
           last_day: "Last 24 hours"
           last_week: "Last week"
@@ -404,6 +406,10 @@ en:
         enabled_chat_bot: "Allow AI bot selector"
         vision_enabled: "Vision enabled"
         ai_bot_user: "AI bot User"
+        cost_input: "Input cost"
+        cost_cached_input: "Cached input cost"
+        cost_output: "Output cost"
+
         save: "Save"
         edit: "Edit"
         saved: "LLM model saved"
@@ -487,6 +493,10 @@ en:
           name: "We include this in the API call to specify which model we'll use"
           vision_enabled: "If enabled, the AI will attempt to understand images. It depends on the model being used supporting vision. Supported by latest models from Anthropic, Google, and OpenAI."
           enabled_chat_bot: "If enabled, users can select this model when creating PMs with the AI bot"
+          cost_input: "The input cost per 1M tokens for this model"
+          cost_cached_input: "The cached input cost per 1M tokens for this model"
+          cost_output: "The output cost per 1M tokens for this model"
+          cost_measure: "$/1M tokens"
         providers:
           aws_bedrock: "AWS Bedrock"
           anthropic: "Anthropic"
diff --git a/db/migrate/20250416215039_add_cost_metrics_to_llm_model.rb b/db/migrate/20250416215039_add_cost_metrics_to_llm_model.rb
new file mode 100644
index 00000000..1df3ccac
--- /dev/null
+++ b/db/migrate/20250416215039_add_cost_metrics_to_llm_model.rb
@@ -0,0 +1,9 @@
+# frozen_string_literal: true
+
+class AddCostMetricsToLlmModel < ActiveRecord::Migration[7.2]
+  def change
+    add_column :llm_models, :input_cost, :float
+    add_column :llm_models, :cached_input_cost, :float
+    add_column :llm_models, :output_cost, :float
+  end
+end
diff --git a/lib/completions/llm.rb b/lib/completions/llm.rb
index a9b67c83..5f82667e 100644
--- a/lib/completions/llm.rb
+++ b/lib/completions/llm.rb
@@ -30,9 +30,26 @@ module DiscourseAi
                       name: "claude-3-7-sonnet",
                       tokens: 200_000,
                       display_name: "Claude 3.7 Sonnet",
+                      input_cost: 3,
+                      cached_input_cost: 0.30,
+                      output_cost: 15,
+                    },
+                    {
+                      name: "claude-3-5-haiku",
+                      tokens: 200_000,
+                      display_name: "Claude 3.5 Haiku",
+                      input_cost: 0.80,
+                      cached_input_cost: 0.08,
+                      output_cost: 4,
+                    },
+                    {
+                      name: "claude-3-opus",
+                      tokens: 200_000,
+                      display_name: "Claude 3 Opus",
+                      input_cost: 15,
+                      cached_input_cost: 1.50,
+                      output_cost: 75,
                     },
-                    { name: "claude-3-5-haiku", tokens: 200_000, display_name: "Claude 3.5 Haiku" },
-                    { name: "claude-3-opus", tokens: 200_000, display_name: "Claude 3 Opus" },
                   ],
                   tokenizer: DiscourseAi::Tokenizer::AnthropicTokenizer,
                   endpoint: "https://api.anthropic.com/v1/messages",
@@ -61,6 +78,8 @@ module DiscourseAi
                       endpoint:
                         "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash-lite",
                       display_name: "Gemini 2.0 Flash Lite",
+                      input_cost: 0.075,
+                      output_cost: 0.30,
                     },
                   ],
                   tokenizer: DiscourseAi::Tokenizer::GeminiTokenizer,
@@ -69,11 +88,46 @@ module DiscourseAi
                 {
                   id: "open_ai",
                   models: [
-                    { name: "o3-mini", tokens: 200_000, display_name: "o3 Mini" },
-                    { name: "o1", tokens: 200_000, display_name: "o1" },
-                    { name: "gpt-4.1", tokens: 800_000, display_name: "GPT-4.1" },
-                    { name: "gpt-4.1-mini", tokens: 800_000, display_name: "GPT-4.1 Mini" },
-                    { name: "gpt-4.1-nano", tokens: 800_000, display_name: "GPT-4.1 Nano" },
+                    {
+                      name: "o3-mini",
+                      tokens: 200_000,
+                      display_name: "o3 Mini",
+                      input_cost: 1.10,
+                      cached_input_cost: 0.55,
+                      output_cost: 4.40,
+                    },
+                    {
+                      name: "o1",
+                      tokens: 200_000,
+                      display_name: "o1",
+                      input_cost: 15,
+                      cached_input_cost: 7.50,
+                      output_cost: 60,
+                    },
+                    {
+                      name: "gpt-4.1",
+                      tokens: 800_000,
+                      display_name: "GPT-4.1",
+                      input_cost: 2,
+                      cached_input_cost: 0.5,
+                      output_cost: 8,
+                    },
+                    {
+                      name: "gpt-4.1-mini",
+                      tokens: 800_000,
+                      display_name: "GPT-4.1 Mini",
+                      input_cost: 0.40,
+                      cached_input_cost: 0.10,
+                      output_cost: 1.60,
+                    },
+                    {
+                      name: "gpt-4.1-nano",
+                      tokens: 800_000,
+                      display_name: "GPT-4.1 Nano",
+                      input_cost: 0.10,
+                      cached_input_cost: 0.025,
+                      output_cost: 0.40,
+                    },
                   ],
                   tokenizer: DiscourseAi::Tokenizer::OpenAiTokenizer,
                   endpoint: "https://api.openai.com/v1/chat/completions",
@@ -86,11 +140,15 @@ module DiscourseAi
                       name: "Meta-Llama-3.3-70B-Instruct",
                       tokens: 131_072,
                       display_name: "Llama 3.3 70B",
+                      input_cost: 0.60,
+                      output_cost: 1.20,
                     },
                     {
                       name: "Meta-Llama-3.1-8B-Instruct",
                       tokens: 16_384,
                       display_name: "Llama 3.1 8B",
+                      input_cost: 0.1,
+                      output_cost: 0.20,
                     },
                   ],
                   tokenizer: DiscourseAi::Tokenizer::Llama3Tokenizer,
diff --git a/lib/completions/report.rb b/lib/completions/report.rb
index 6982b9a7..7290a6d3 100644
--- a/lib/completions/report.rb
+++ b/lib/completions/report.rb
@@ -33,6 +33,27 @@ module DiscourseAi
         stats.total_requests || 0
       end
 
+      def total_spending
+        total = total_input_spending + total_output_spending + total_cached_input_spending
+        total.round(2)
+      end
+
+      def total_input_spending
+        model_costs.sum { |row| row.input_cost.to_f * row.total_request_tokens.to_i / 1_000_000.0 }
+      end
+
+      def total_output_spending
+        model_costs.sum do |row|
+          row.output_cost.to_f * row.total_response_tokens.to_i / 1_000_000.0
+        end
+      end
+
+      def total_cached_input_spending
+        model_costs.sum do |row|
+          row.cached_input_cost.to_f * row.total_cached_tokens.to_i / 1_000_000.0
+        end
+      end
+
       def stats
         @stats ||=
           base_query.select(
@@ -46,6 +67,24 @@ module DiscourseAi
           ]
       end
 
+      def model_costs
+        @model_costs ||=
+          base_query
+            .joins("LEFT JOIN llm_models ON llm_models.name = language_model")
+            .group(
+              "llm_models.name, llm_models.input_cost, llm_models.output_cost, llm_models.cached_input_cost",
+            )
+            .select(
+              "llm_models.name",
+              "llm_models.input_cost",
+              "llm_models.output_cost",
+              "llm_models.cached_input_cost",
+              "SUM(COALESCE(request_tokens, 0)) as total_request_tokens",
+              "SUM(COALESCE(response_tokens, 0)) as total_response_tokens",
+              "SUM(COALESCE(cached_tokens, 0)) as total_cached_tokens",
+            )
+      end
+
       def guess_period(period = nil)
         period = nil if %i[day month hour].include?(period)
         period ||
@@ -76,7 +115,15 @@ module DiscourseAi
       def user_breakdown
         base_query
           .joins(:user)
-          .group(:user_id, "users.username", "users.uploaded_avatar_id")
+          .joins("LEFT JOIN llm_models ON llm_models.name = language_model")
+          .group(
+            :user_id,
+            "users.username",
+            "users.uploaded_avatar_id",
+            "llm_models.input_cost",
+            "llm_models.output_cost",
+            "llm_models.cached_input_cost",
+          )
           .order("usage_count DESC")
           .limit(USER_LIMIT)
           .select(
@@ -87,12 +134,21 @@ module DiscourseAi
             "SUM(COALESCE(cached_tokens,0)) as total_cached_tokens",
             "SUM(COALESCE(request_tokens,0)) as total_request_tokens",
             "SUM(COALESCE(response_tokens,0)) as total_response_tokens",
+            "SUM(COALESCE(request_tokens, 0)) * COALESCE(llm_models.input_cost, 0) / 1000000.0 as input_spending",
+            "SUM(COALESCE(response_tokens, 0)) * COALESCE(llm_models.output_cost, 0) / 1000000.0 as output_spending",
+            "SUM(COALESCE(cached_tokens, 0)) * COALESCE(llm_models.cached_input_cost, 0) / 1000000.0 as cached_input_spending",
           )
       end
 
       def feature_breakdown
         base_query
-          .group(:feature_name)
+          .joins("LEFT JOIN llm_models ON llm_models.name = language_model")
+          .group(
+            :feature_name,
+            "llm_models.input_cost",
+            "llm_models.output_cost",
+            "llm_models.cached_input_cost",
+          )
           .order("usage_count DESC")
           .select(
             "case when coalesce(feature_name, '') = '' then '#{UNKNOWN_FEATURE}' else feature_name end as feature_name",
@@ -101,12 +157,21 @@ module DiscourseAi
             "SUM(COALESCE(cached_tokens,0)) as total_cached_tokens",
             "SUM(COALESCE(request_tokens,0)) as total_request_tokens",
             "SUM(COALESCE(response_tokens,0)) as total_response_tokens",
+            "SUM(COALESCE(request_tokens, 0)) * COALESCE(llm_models.input_cost, 0) / 1000000.0 as input_spending",
+            "SUM(COALESCE(response_tokens, 0)) * COALESCE(llm_models.output_cost, 0) / 1000000.0 as output_spending",
+            "SUM(COALESCE(cached_tokens, 0)) * COALESCE(llm_models.cached_input_cost, 0) / 1000000.0 as cached_input_spending",
           )
       end
 
       def model_breakdown
         base_query
-          .group(:language_model)
+          .joins("LEFT JOIN llm_models ON llm_models.name = language_model")
+          .group(
+            :language_model,
+            "llm_models.input_cost",
+            "llm_models.output_cost",
+            "llm_models.cached_input_cost",
+          )
           .order("usage_count DESC")
           .select(
             "language_model as llm",
@@ -115,6 +180,9 @@ module DiscourseAi
             "SUM(COALESCE(cached_tokens,0)) as total_cached_tokens",
             "SUM(COALESCE(request_tokens,0)) as total_request_tokens",
             "SUM(COALESCE(response_tokens,0)) as total_response_tokens",
+            "SUM(COALESCE(request_tokens, 0)) * COALESCE(llm_models.input_cost, 0) / 1000000.0 as input_spending",
+            "SUM(COALESCE(response_tokens, 0)) * COALESCE(llm_models.output_cost, 0) / 1000000.0 as output_spending",
+            "SUM(COALESCE(cached_tokens, 0)) * COALESCE(llm_models.cached_input_cost, 0) / 1000000.0 as cached_input_spending",
           )
       end
 
diff --git a/spec/fabricators/llm_model_fabricator.rb b/spec/fabricators/llm_model_fabricator.rb
index 87b1ef25..d61bb915 100644
--- a/spec/fabricators/llm_model_fabricator.rb
+++ b/spec/fabricators/llm_model_fabricator.rb
@@ -8,6 +8,9 @@ Fabricator(:llm_model) do
   api_key "123"
   url "https://api.openai.com/v1/chat/completions"
   max_prompt_tokens 131_072
+  input_cost 10
+  cached_input_cost 2.5
+  output_cost 40
 end
 
 Fabricator(:anthropic_model, from: :llm_model) do
diff --git a/spec/requests/admin/ai_usage_controller_spec.rb b/spec/requests/admin/ai_usage_controller_spec.rb
index eae9285b..f411ed41 100644
--- a/spec/requests/admin/ai_usage_controller_spec.rb
+++ b/spec/requests/admin/ai_usage_controller_spec.rb
@@ -5,6 +5,7 @@ require "rails_helper"
 RSpec.describe DiscourseAi::Admin::AiUsageController do
   fab!(:admin)
   fab!(:user)
+  fab!(:llm_model)
   let(:usage_report_path) { "/admin/plugins/discourse-ai/ai-usage-report.json" }
 
   before { SiteSetting.discourse_ai_enabled = true }
@@ -35,6 +36,18 @@ RSpec.describe DiscourseAi::Admin::AiUsageController do
         )
       end
 
+      fab!(:log3) do
+        AiApiAuditLog.create!(
+          provider_id: 1,
+          feature_name: "ai_helper",
+          language_model: llm_model.name,
+          request_tokens: 300,
+          response_tokens: 150,
+          cached_tokens: 50,
+          created_at: 3.days.ago,
+        )
+      end
+
       it "returns correct data structure" do
         get usage_report_path
 
@@ -55,7 +68,7 @@ RSpec.describe DiscourseAi::Admin::AiUsageController do
             }
 
         json = response.parsed_body
-        expect(json["summary"]["total_tokens"]).to eq(450) # sum of all tokens
+        expect(json["summary"]["total_tokens"]).to eq(900) # sum of all tokens
       end
 
       it "filters by feature" do
@@ -79,6 +92,26 @@ RSpec.describe DiscourseAi::Admin::AiUsageController do
         expect(models.first["total_tokens"]).to eq(300)
       end
 
+      it "shows an estimated cost" do
+        get usage_report_path, params: { model: llm_model.name }
+
+        json = response.parsed_body
+        summary = json["summary"]
+        feature = json["features"].find { |f| f["feature_name"] == "ai_helper" }
+
+        expected_input_spending = llm_model.input_cost * log3.request_tokens / 1_000_000.0
+        expected_cached_input_spending =
+          llm_model.cached_input_cost * log3.cached_tokens / 1_000_000.0
+        expected_output_spending = llm_model.output_cost * log3.response_tokens / 1_000_000.0
+        expected_total_spending =
+          expected_input_spending + expected_cached_input_spending + expected_output_spending
+
+        expect(feature["input_spending"].to_s).to eq(expected_input_spending.to_s)
+        expect(feature["output_spending"].to_s).to eq(expected_output_spending.to_s)
+        expect(feature["cached_input_spending"].to_s).to eq(expected_cached_input_spending.to_s)
+        expect(summary["total_spending"].to_s).to eq(expected_total_spending.round(2).to_s)
+      end
+
       it "handles different period groupings" do
         get usage_report_path, params: { period: "hour" }
         expect(response.status).to eq(200)