diff --git a/model/textparse/openmetricsparse.go b/model/textparse/openmetricsparse.go
index b7ad1dd85c..43708d0dcc 100644
--- a/model/textparse/openmetricsparse.go
+++ b/model/textparse/openmetricsparse.go
@@ -222,9 +222,82 @@ func (p *OpenMetricsParser) Exemplar(e *exemplar.Exemplar) bool {
 // CreatedTimestamp returns nil as it's not implemented yet.
 // TODO(bwplotka): https://github.com/prometheus/prometheus/issues/12980
 func (p *OpenMetricsParser) CreatedTimestamp() *int64 {
-	return nil
-}
+	newParser := deepCopyParser(p)
+	switch t, _ := newParser.Next(); t {
+		case EntrySeries:
+			for {
+				// Check _created suffix
+				var lbs labels.Labels
+				newParser.Metric(&lbs)
+				name := lbs.Get(model.MetricNameLabel)
+				if name[len(name)-8:] != "_created" {
+					return nil
+				}
 
+				var labelsPrev labels.Labels
+				p.Metric(&labelsPrev) 
+				namePrev := labelsPrev.Get(model.MetricNameLabel)
+				// a case like foo_total != foo_created fails even if underlying metric is the same
+				if namePrev != name { 
+					return nil
+				}
+
+
+				// TODO: for histograms
+				if t, _ := newParser.Next(); t != EntrySeries {
+					return nil
+				}
+			}
+		default:
+			// If not series, we don't care.
+		}
+	return nil
+ }
+
+ // We need this because we want offsets of the original parser unchanged when 
+ // we're working with a new parser in CreatedTimeStamp()
+ func deepCopyParser(p *OpenMetricsParser) OpenMetricsParser {
+	newB := make([]byte, len(p.l.b))
+	copy(newB, p.l.b)
+
+	newLexer := &openMetricsLexer{
+		b:    newB,
+		i:    p.l.i,
+		start: p.l.start,
+		err:  p.l.err,
+		state: p.l.state,
+	}
+
+	newSeries := make([]byte, len(p.series))
+	copy(newSeries, p.series)
+	newText := make([]byte, len(p.text))
+	copy(newText, p.text)
+	newOffsets := make([]int, len(p.offsets))
+	copy(newOffsets, p.offsets)
+	newEOffsets := p.eOffsets
+	newExemplar := p.exemplar
+	newParser := OpenMetricsParser{
+		l       :newLexer,
+		builder :p.builder,
+		series  :newSeries,
+		text    :newText,
+		mtype   :p.mtype,
+		val     :p.val,
+		ts      :p.ts,
+		hasTS   :p.hasTS,
+		start   :p.start,
+
+		offsets :newOffsets,
+	
+		eOffsets      :newEOffsets,
+		exemplar      :newExemplar,
+		exemplarVal   :p.exemplarVal,
+		exemplarTs    :p.exemplarTs,
+		hasExemplarTs :p.hasExemplarTs,
+	}
+	return newParser
+ }
+ 
 // nextToken returns the next token from the openMetricsLexer.
 func (p *OpenMetricsParser) nextToken() token {
 	tok := p.l.Lex()
diff --git a/model/textparse/openmetricsparse_test.go b/model/textparse/openmetricsparse_test.go
index bc76a540d3..6334239922 100644
--- a/model/textparse/openmetricsparse_test.go
+++ b/model/textparse/openmetricsparse_test.go
@@ -64,7 +64,10 @@ _metric_starting_with_underscore 1
 testmetric{_label_starting_with_underscore="foo"} 1
 testmetric{label="\"bar\""} 1
 # TYPE foo counter
-foo_total 17.0 1520879607.789 # {id="counter-test"} 5`
+foo_total 17.0 1520879607.789 # {id="counter-test"} 5
+foo_created 1000
+foo_total{a="b"} 17.0 1520879607.789 # {id="counter-test"} 5
+foo_created{a="b"} 1000`
 
 	input += "\n# HELP metric foo\x00bar"
 	input += "\nnull_byte_metric{a=\"abc\x00\"} 1"
@@ -225,6 +228,24 @@ foo_total 17.0 1520879607.789 # {id="counter-test"} 5`
 			lset: labels.FromStrings("__name__", "foo_total"),
 			t:    int64p(1520879607789),
 			e:    &exemplar.Exemplar{Labels: labels.FromStrings("id", "counter-test"), Value: 5},
+			ct: int64p(1000),
+		},{
+			m: `foo_total{a="b"}`,
+			v: 17.0,
+			lset: labels.FromStrings("__name__", "foo_total", "a", "b"),
+			t: int64p(1520879607789),
+			e: &exemplar.Exemplar{Labels: labels.FromStrings("id", "counter-test"), Value: 5},
+			ct: int64p(1000),
+		},{
+			m: "rpc_durations_histogram_seconds_bucket",
+			ct: int64p(1000),
+		},
+		{
+			m:    "foo_total",
+			v:    17,
+			lset: labels.FromStrings("__name__", "foo_total"),
+			t:    int64p(1520879607789),
+			e:    &exemplar.Exemplar{Labels: labels.FromStrings("id", "counter-test"), Value: 5},
 		}, {
 			m:    "metric",
 			help: "foo\x00bar",
@@ -683,4 +704,4 @@ func TestOMNullByteHandling(t *testing.T) {
 
 		require.Equal(t, c.err, err.Error(), "test %d", i)
 	}
-}
+}
\ No newline at end of file
diff --git a/model/textparse/promparse_test.go b/model/textparse/promparse_test.go
index 66986291d7..93906ec042 100644
--- a/model/textparse/promparse_test.go
+++ b/model/textparse/promparse_test.go
@@ -41,6 +41,7 @@ type expectedParse struct {
 	unit    string
 	comment string
 	e       *exemplar.Exemplar
+	ct 	*int64
 }
 
 func TestPromParse(t *testing.T) {
@@ -218,6 +219,11 @@ func checkParseResults(t *testing.T, p Parser, exp []expectedParse) {
 				require.True(t, found)
 				testutil.RequireEqual(t, *exp[i].e, e)
 			}
+			if ct := p.CreatedTimestamp(); ct != nil {
+				require.Equal(t, *exp[i].ct, *ct)
+			} else {
+				require.Nil(t, exp[i].ct)
+			}
 
 		case EntryType:
 			m, typ := p.Type()