diff --git a/package.json b/package.json
index 9db7c2905..92c16d58b 100644
--- a/package.json
+++ b/package.json
@@ -95,11 +95,11 @@
         "promptpex:test-st-def:ollama": "genaiscript run promptpex \"samples/speech-tag/speech-tag.prompty\"  --vars \"effort=min\"  --vars \"out=evals/test-st-def\" --env .env.ollama",
         "promptpex:test-st-mingt": "genaiscript run promptpex \"samples/speech-tag/speech-tag.prompty\"  --vars \"effort=min\" --vars \"groundtruthModel=azure:gpt-4.1-mini_2025-04-14\" --vars \"evalModel=ollama:llama3.3\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"modelsUnderTest=ollama:llama3.3\" --vars \"out=evals/test-st-mingt\"",
         "promptpex:test-st-mediumgt": "genaiscript run promptpex \"samples/speech-tag/speech-tag-multi.prompty\"  --vars \"effort=medium\" --vars \"groundtruthModel=azure:gpt-4.1-mini_2025-04-14\" --vars \"evalModel=ollama:llama3.3\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"modelsUnderTest=azure:gpt-4.1-mini_2025-04-14;ollama:llama3.3\" --vars \"out=evals/test-st-mediumgt\"",
-        "promptpex:test-st-min-gen": "genaiscript run promptpex \"samples/speech-tag/speech-tag.prompty\"  --vars \"effort=min\" --vars \"groundtruthModel=azure:gpt-4.1-mini_2025-04-14\" --vars \"evals=false\" --vars \"compliance=false\" --vars \"baselineTests=false\" --vars \"evalModelGroundtruth=azure:gpt-4.1-mini_2025-04-14;ollama:llama3.3\" --vars \"out=evals/test-st-min-gen\" --env .env.ollama",
-        "promptpex:test-st-min-run": "genaiscript run promptpex \"evals/test-st-min-gen/speech-tag/promptpex_context.json\" --vars \"evals=false\" --vars \"compliance=false\" --vars \"baselineTests=false\" --vars \"modelsUnderTest=ollama:qwen2.5:3b;ollama:llama3.2:1b;ollama:llama3.3\" --vars \"out=evals/test-st-min-run\"",
-        "promptpex:test-st-min-eval": "genaiscript run promptpex \"evals/test-st-min-run/speech-tag/promptpex_context.json\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"evalModel=azure:gpt-4.1-mini_2025-04-14\" --vars \"out=evals/test-st-min-eval\"",
-        "promptpex:test-st-min-eval1": "genaiscript run promptpex \"evals/test-st-min-run/speech-tag/promptpex_context.json\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"evalModel=azure:gpt-4.1-mini_2025-04-14;ollama:llama3.3\" --vars \"out=evals/test-st-min-eval\"",
-        "promptpex:test-st-min-runeval": "genaiscript run promptpex \"evals/test-st-min-gen/speech-tag/promptpex_context.json\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"modelsUnderTest=ollama:qwen2.5:3b;ollama:llama3.2:1b\" --vars \"evalModel=ollama:llama3.3\" --vars \"out=evals/test-st-min-runeval\"",
+        "promptpex:test-st-min-gen:ollama": "genaiscript run promptpex \"samples/speech-tag/speech-tag.prompty\"  --vars \"effort=min\" --vars \"groundtruthModel=azure:gpt-4.1-mini_2025-04-14\" --vars \"evals=false\" --vars \"compliance=false\" --vars \"baselineTests=false\" --vars \"evalModelGroundtruth=azure:gpt-4.1-mini_2025-04-14;ollama:llama3.3\" --vars \"out=evals/test-st-min-gen\" --env .env.ollama",
+        "promptpex:test-st-min-run:ollama": "genaiscript run promptpex \"evals/test-st-min-gen/speech-tag/promptpex_context.json\" --vars \"evals=false\" --vars \"compliance=false\" --vars \"baselineTests=false\" --vars \"modelsUnderTest=ollama:qwen2.5:3b;ollama:llama3.2:1b;ollama:llama3.3\" --vars \"out=evals/test-st-min-run --env .env.ollama\"",
+        "promptpex:test-st-min-eval:ollama": "genaiscript run promptpex \"evals/test-st-min-run/speech-tag/promptpex_context.json\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"evalModel=azure:gpt-4.1-mini_2025-04-14\" --vars \"out=evals/test-st-min-eval\" --env .env.ollama",
+        "promptpex:test-st-min-eval1": "genaiscript run promptpex \"evals/test-st-min-run/speech-tag/promptpex_context.json\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"evalModel=azure:gpt-4.1-mini_2025-04-14;ollama:llama3.3\" --vars \"out=evals/test-st-min-eval\" --env .env.ollama",
+        "promptpex:test-st-min-runeval:ollama": "genaiscript run promptpex \"evals/test-st-min-gen/speech-tag/promptpex_context.json\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\" --vars \"modelsUnderTest=ollama:qwen2.5:3b;ollama:llama3.2:1b\" --vars \"evalModel=ollama:llama3.3\" --vars \"out=evals/test-st-min-runeval\" --env .env.ollama",
         "promptpex:test-st-med": "genaiscript run promptpex \"samples/speech-tag/speech-tag.prompty\"  --vars \"effort=medium\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\"  --vars \"modelsUnderTest=ollama:llama3.2:1b\" --vars \"out=evals/test-st-med\"",
         "promptpex:test-headline": "genaiscript run promptpex \"samples/demo/rate-headline.prompty\"  --vars \"effort=min\" --vars \"evals=true\" --vars \"compliance=true\" --vars \"baselineTests=false\"  --vars \"modelsUnderTest=ollama:llama3.2:1b\" --vars \"out=evals/test-headline\"",
         "promptpex:test1": "genaiscript run promptpex \"samples/demo/rate-headline.prompty\"  --vars \"cache=true\" --vars \"evals=false\" --vars \"effort=min\" --vars \"baselineTests=true\"  --vars \"filterTestCount=3\"  --vars \"modelsUnderTest=ollama:llama3.2:1b\" --vars \"out=evals/test-test1\"",
diff --git a/src/genaisrc/src/loaders.mts b/src/genaisrc/src/loaders.mts
index ee9b7d037..dff011263 100644
--- a/src/genaisrc/src/loaders.mts
+++ b/src/genaisrc/src/loaders.mts
@@ -112,6 +112,7 @@ export async function loadPromptContext(
     let rateTests = path.join(dir, "test_collection_review.md")
     let testData = path.join(dir, "test_data.json")
     let testResults = path.join(dir, "test_results.json")
+    let groundtruthResults = path.join(dir, "test_groundtruth.json")
     let testEvals = path.join(dir, "test_evals.json")
     let baselineTestEvals = path.join(dir, "baseline_test_evals.json")
     let ruleEvals = path.join(dir, "rule_evals.json")
@@ -193,6 +194,7 @@ export async function loadPromptContext(
         prompt: promptFile,
         originalPrompt: originalPromptFile,
         testOutputs: await workspace.readText(testResults),
+        groundtruthOutputs: await workspace.readText(groundtruthResults),
         intent: await workspace.readText(intent),
         inputSpec: await workspace.readText(inputSpec),
         rules: tidyRulesFile(await workspace.readText(rules)),
@@ -372,6 +374,7 @@ async function loadPromptContextFromJSON(
         ctx.rateTests,
         ctx.testData,
         ctx.testOutputs,
+        ctx.groundtruthOutputs,
         ctx.testEvals,
         ctx.baselineTestEvals,
         ctx.ruleEvals,
diff --git a/src/genaisrc/src/parsers.mts b/src/genaisrc/src/parsers.mts
index 6853212cd..590b02140 100644
--- a/src/genaisrc/src/parsers.mts
+++ b/src/genaisrc/src/parsers.mts
@@ -139,6 +139,8 @@ export function parseTestResults(
     const rules = parseRules(files.rules.content)
     const res = (parsers.JSON5(files.testOutputs.content) ||
         []) as PromptPexTestResult[]
+    const groundtruthRes = (parsers.JSON5(files.groundtruthOutputs.content) ||
+        []) as PromptPexTestResult[]
     res.forEach((r) => {
         r.inverse =
             r.ruleid !== null && parseInt(r.ruleid as any) > rules.length
@@ -151,6 +153,13 @@ export function parseTestResults(
         if (diagnostics)
             throw new Error(`missing 'model' for test result ${r.id}`)
     }
+    for (const r of groundtruthRes.filter((r) => !r.error && !r.model)) {
+        output.warn(
+            `missing 'model' for groundtruth test result ${r.id} in ${files.groundtruthOutputs.filename}`
+        )
+        if (diagnostics)
+            throw new Error(`missing 'model' for groundtruth test result ${r.id}`)
+    }
     for (const r of res) if (isNaN(r.ruleid)) r.ruleid = null
     return res
 }
diff --git a/src/genaisrc/src/promptpex.mts b/src/genaisrc/src/promptpex.mts
index 5a5a66aff..3e6d8c910 100644
--- a/src/genaisrc/src/promptpex.mts
+++ b/src/genaisrc/src/promptpex.mts
@@ -254,10 +254,10 @@ export async function promptpexGenerate(files: PromptPexContext) {
                 renderTestResults(results.filter((r) => r.isGroundtruth)),
                 { maxRows: 12 }
             )
-            files.testOutputs.content = JSON.stringify(results, null, 2)
+            files.groundtruthOutputs.content = JSON.stringify(results, null, 2)
             if (files.writeResults)
                 await workspace.writeText(
-                    files.testOutputs.filename,
+                    files.groundtruthOutputs.filename,
                     JSON.stringify(results, null, 2)
                 )
         } else {
diff --git a/src/genaisrc/src/types.mts b/src/genaisrc/src/types.mts
index 9a6441b3a..9ed3bcb3a 100644
--- a/src/genaisrc/src/types.mts
+++ b/src/genaisrc/src/types.mts
@@ -284,6 +284,11 @@ export interface PromptPexContext {
      */
     testOutputs: WorkspaceFile
 
+    /**
+     * Groundtruth Output (TO) - Test results with metrics for groundtruth test outputs
+     */
+    groundtruthOutputs: WorkspaceFile
+
     /**
      * Coverage and validate test evals
      */