diff --git a/app/models/ai_api_audit_log.rb b/app/models/ai_api_audit_log.rb
new file mode 100644
index 00000000..c3718574
--- /dev/null
+++ b/app/models/ai_api_audit_log.rb
@@ -0,0 +1,7 @@
+# frozen_string_literal: true
+
+class AiApiAuditLog < ActiveRecord::Base
+  module Provider
+    OpenAI = 1
+  end
+end
diff --git a/db/migrate/20230424055354_create_ai_api_audit_logs.rb b/db/migrate/20230424055354_create_ai_api_audit_logs.rb
new file mode 100644
index 00000000..eb4c6413
--- /dev/null
+++ b/db/migrate/20230424055354_create_ai_api_audit_logs.rb
@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+class CreateAiApiAuditLogs < ActiveRecord::Migration[7.0]
+  def change
+    create_table :ai_api_audit_logs do |t|
+      t.integer :provider_id, null: false
+      t.integer :user_id
+      t.integer :request_tokens
+      t.integer :response_tokens
+      t.string :raw_request_payload
+      t.string :raw_response_payload
+      t.timestamps
+    end
+  end
+end
diff --git a/lib/shared/inference/openai_completions.rb b/lib/shared/inference/openai_completions.rb
index c7553c8f..63643e16 100644
--- a/lib/shared/inference/openai_completions.rb
+++ b/lib/shared/inference/openai_completions.rb
@@ -14,6 +14,7 @@ module ::DiscourseAi
         top_p: nil,
         max_tokens: nil,
         stream: false,
+        user_id: nil,
         &blk
       )
         raise ArgumentError, "block must be supplied in streaming mode" if stream && !blk
@@ -39,7 +40,8 @@ module ::DiscourseAi
           write_timeout: TIMEOUT,
         ) do |http|
           request = Net::HTTP::Post.new(url, headers)
-          request.body = payload.to_json
+          request_body = payload.to_json
+          request.body = request_body
 
           response = http.request(request)
 
@@ -50,24 +52,44 @@ module ::DiscourseAi
             raise CompletionFailed
           end
 
+          log =
+            AiApiAuditLog.create!(
+              provider_id: AiApiAuditLog::Provider::OpenAI,
+              raw_request_payload: request_body,
+              user_id: user_id,
+            )
+
           if stream
-            stream(http, response, &blk)
+            stream(http, response, messages, log, &blk)
           else
-            JSON.parse(response.read_body, symbolize_names: true)
+            response_body = response.body
+            parsed = JSON.parse(response_body, symbolize_names: true)
+
+            log.update!(
+              raw_response_payload: response_body,
+              request_tokens: parsed.dig(:usage, :prompt_tokens),
+              response_tokens: parsed.dig(:usage, :completion_tokens),
+            )
+            parsed
           end
         end
       end
 
-      def self.stream(http, response)
+      def self.stream(http, response, messages, log)
         cancelled = false
         cancel = lambda { cancelled = true }
 
+        response_data = +""
+        response_raw = +""
+
         response.read_body do |chunk|
           if cancelled
             http.finish
             break
           end
 
+          response_raw << chunk
+
           chunk
             .split("\n")
             .each do |line|
@@ -75,7 +97,15 @@ module ::DiscourseAi
 
               next if data == "[DONE]"
 
-              yield JSON.parse(data, symbolize_names: true), cancel if data
+              if data
+                json = JSON.parse(data, symbolize_names: true)
+                choices = json[:choices]
+                if choices && choices[0]
+                  delta = choices[0].dig(:delta, :content)
+                  response_data << delta if delta
+                end
+                yield json, cancel
+              end
 
               if cancelled
                 http.finish
@@ -85,6 +115,16 @@ module ::DiscourseAi
         end
       rescue IOError
         raise if !cancelled
+      ensure
+        log.update!(
+          raw_response_payload: response_raw,
+          request_tokens: DiscourseAi::Tokenizer.size(extract_prompt(messages)),
+          response_tokens: DiscourseAi::Tokenizer.size(response_data),
+        )
+      end
+
+      def self.extract_prompt(messages)
+        messages.map { |message| message[:content] || message["content"] || "" }.join("\n")
       end
     end
   end
diff --git a/spec/shared/inference/openai_completions_spec.rb b/spec/shared/inference/openai_completions_spec.rb
index 4624b1fd..cac34c22 100644
--- a/spec/shared/inference/openai_completions_spec.rb
+++ b/spec/shared/inference/openai_completions_spec.rb
@@ -18,6 +18,8 @@ describe DiscourseAi::Inference::OpenAiCompletions do
       },
     ).to_return(status: 200, body: body, headers: {})
 
+    user_id = 183
+
     prompt = [role: "user", content: "write 3 words"]
     completions =
       DiscourseAi::Inference::OpenAiCompletions.perform!(
@@ -26,10 +28,24 @@ describe DiscourseAi::Inference::OpenAiCompletions do
         temperature: 0.5,
         top_p: 0.8,
         max_tokens: 700,
+        user_id: user_id,
       )
     expect(completions[:choices][0][:message][:content]).to eq(
       "1. Serenity\n2. Laughter\n3. Adventure",
     )
+
+    expect(AiApiAuditLog.count).to eq(1)
+    log = AiApiAuditLog.first
+
+    request_body = (<<~JSON).strip
+      {"model":"gpt-3.5-turbo","messages":[{"role":"user","content":"write 3 words"}],"temperature":0.5,"top_p":0.8,"max_tokens":700}
+    JSON
+
+    expect(log.provider_id).to eq(AiApiAuditLog::Provider::OpenAI)
+    expect(log.request_tokens).to eq(12)
+    expect(log.response_tokens).to eq(13)
+    expect(log.raw_request_payload).to eq(request_body)
+    expect(log.raw_response_payload).to eq(body)
   end
 
   it "raises an error if attempting to stream without a block" do
@@ -88,5 +104,18 @@ describe DiscourseAi::Inference::OpenAiCompletions do
     end
 
     expect(content).to eq("Mountain Tree ")
+
+    expect(AiApiAuditLog.count).to eq(1)
+    log = AiApiAuditLog.first
+
+    request_body = (<<~JSON).strip
+      {"model":"gpt-3.5-turbo","messages":[{"role":"user","content":"write 3 words"}],"stream":true}
+    JSON
+
+    expect(log.provider_id).to eq(AiApiAuditLog::Provider::OpenAI)
+    expect(log.request_tokens).to eq(5)
+    expect(log.response_tokens).to eq(4)
+    expect(log.raw_request_payload).to eq(request_body)
+    expect(log.raw_response_payload).to eq(payload)
   end
 end