@botpress/cognitive 0.4.7 → 0.5.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.turbo/turbo-build.log +7 -7
- package/dist/index.cjs +784 -111
- package/dist/index.cjs.map +2 -2
- package/dist/index.d.ts +177 -112
- package/dist/index.mjs +784 -111
- package/dist/index.mjs.map +2 -2
- package/package.json +1 -1
- package/refresh-models.ts +51 -30
package/dist/index.mjs
CHANGED
|
@@ -646,10 +646,79 @@ var models = {
|
|
|
646
646
|
tags: ["recommended", "reasoning", "general-purpose", "vision", "coding", "agents"],
|
|
647
647
|
lifecycle: "production",
|
|
648
648
|
capabilities: {
|
|
649
|
-
supportsImages: true
|
|
649
|
+
supportsImages: true,
|
|
650
|
+
supportsAudio: false,
|
|
651
|
+
supportsTranscription: false,
|
|
652
|
+
supportsSearch: true
|
|
650
653
|
},
|
|
651
654
|
aliases: ["gpt-5.4"]
|
|
652
655
|
},
|
|
656
|
+
"openai:gpt-5.4-mini-2026-03-17": {
|
|
657
|
+
id: "openai:gpt-5.4-mini-2026-03-17",
|
|
658
|
+
name: "GPT-5.4 Mini",
|
|
659
|
+
description: "GPT-5.4 Mini brings the strengths of GPT-5.4 to a faster, more efficient model designed for high-volume workloads. It is optimized for speed and cost while retaining strong reasoning and vision capabilities.",
|
|
660
|
+
input: {
|
|
661
|
+
maxTokens: 4e5,
|
|
662
|
+
costPer1MTokens: 0.75
|
|
663
|
+
},
|
|
664
|
+
output: {
|
|
665
|
+
maxTokens: 128e3,
|
|
666
|
+
costPer1MTokens: 4.5
|
|
667
|
+
},
|
|
668
|
+
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
669
|
+
lifecycle: "production",
|
|
670
|
+
capabilities: {
|
|
671
|
+
supportsImages: true,
|
|
672
|
+
supportsAudio: false,
|
|
673
|
+
supportsTranscription: false,
|
|
674
|
+
supportsSearch: false
|
|
675
|
+
},
|
|
676
|
+
aliases: ["gpt-5.4-mini"]
|
|
677
|
+
},
|
|
678
|
+
"openai:gpt-5.4-nano-2026-03-17": {
|
|
679
|
+
id: "openai:gpt-5.4-nano-2026-03-17",
|
|
680
|
+
name: "GPT-5.4 Nano",
|
|
681
|
+
description: "GPT-5.4 Nano is the smallest and cheapest GPT-5.4 variant, designed for tasks where speed and cost matter most like classification, data extraction, ranking, and coding sub-agents.",
|
|
682
|
+
input: {
|
|
683
|
+
maxTokens: 4e5,
|
|
684
|
+
costPer1MTokens: 0.2
|
|
685
|
+
},
|
|
686
|
+
output: {
|
|
687
|
+
maxTokens: 128e3,
|
|
688
|
+
costPer1MTokens: 1.25
|
|
689
|
+
},
|
|
690
|
+
tags: ["low-cost", "reasoning", "general-purpose", "vision"],
|
|
691
|
+
lifecycle: "production",
|
|
692
|
+
capabilities: {
|
|
693
|
+
supportsImages: true,
|
|
694
|
+
supportsAudio: false,
|
|
695
|
+
supportsTranscription: false,
|
|
696
|
+
supportsSearch: false
|
|
697
|
+
},
|
|
698
|
+
aliases: ["gpt-5.4-nano"]
|
|
699
|
+
},
|
|
700
|
+
"openai:gpt-5.3-chat": {
|
|
701
|
+
id: "openai:gpt-5.3-chat",
|
|
702
|
+
name: "GPT-5.3 Chat",
|
|
703
|
+
description: "GPT-5.3 Chat is the GPT-5.3 Instant model used in ChatGPT, exposed via the API. Rolling alias that points to the latest snapshot.",
|
|
704
|
+
input: {
|
|
705
|
+
maxTokens: 128e3,
|
|
706
|
+
costPer1MTokens: 1.75
|
|
707
|
+
},
|
|
708
|
+
output: {
|
|
709
|
+
maxTokens: 16384,
|
|
710
|
+
costPer1MTokens: 14
|
|
711
|
+
},
|
|
712
|
+
tags: ["reasoning", "general-purpose", "vision"],
|
|
713
|
+
lifecycle: "preview",
|
|
714
|
+
capabilities: {
|
|
715
|
+
supportsImages: true,
|
|
716
|
+
supportsAudio: false,
|
|
717
|
+
supportsTranscription: false,
|
|
718
|
+
supportsSearch: false
|
|
719
|
+
},
|
|
720
|
+
aliases: ["gpt-5.3-chat-latest"]
|
|
721
|
+
},
|
|
653
722
|
"openai:gpt-5.2-2025-12-11": {
|
|
654
723
|
id: "openai:gpt-5.2-2025-12-11",
|
|
655
724
|
name: "GPT-5.2",
|
|
@@ -665,7 +734,10 @@ var models = {
|
|
|
665
734
|
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
666
735
|
lifecycle: "production",
|
|
667
736
|
capabilities: {
|
|
668
|
-
supportsImages: true
|
|
737
|
+
supportsImages: true,
|
|
738
|
+
supportsAudio: false,
|
|
739
|
+
supportsTranscription: false,
|
|
740
|
+
supportsSearch: true
|
|
669
741
|
}
|
|
670
742
|
},
|
|
671
743
|
"openai:gpt-5.1-2025-11-13": {
|
|
@@ -683,7 +755,10 @@ var models = {
|
|
|
683
755
|
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
684
756
|
lifecycle: "production",
|
|
685
757
|
capabilities: {
|
|
686
|
-
supportsImages: true
|
|
758
|
+
supportsImages: true,
|
|
759
|
+
supportsAudio: false,
|
|
760
|
+
supportsTranscription: false,
|
|
761
|
+
supportsSearch: true
|
|
687
762
|
}
|
|
688
763
|
},
|
|
689
764
|
"openai:gpt-5-2025-08-07": {
|
|
@@ -701,7 +776,10 @@ var models = {
|
|
|
701
776
|
tags: ["reasoning", "general-purpose", "vision"],
|
|
702
777
|
lifecycle: "production",
|
|
703
778
|
capabilities: {
|
|
704
|
-
supportsImages: true
|
|
779
|
+
supportsImages: true,
|
|
780
|
+
supportsAudio: false,
|
|
781
|
+
supportsTranscription: false,
|
|
782
|
+
supportsSearch: false
|
|
705
783
|
},
|
|
706
784
|
aliases: ["gpt-5"]
|
|
707
785
|
},
|
|
@@ -720,7 +798,10 @@ var models = {
|
|
|
720
798
|
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
721
799
|
lifecycle: "production",
|
|
722
800
|
capabilities: {
|
|
723
|
-
supportsImages: true
|
|
801
|
+
supportsImages: true,
|
|
802
|
+
supportsAudio: false,
|
|
803
|
+
supportsTranscription: false,
|
|
804
|
+
supportsSearch: false
|
|
724
805
|
},
|
|
725
806
|
aliases: ["gpt-5-mini"]
|
|
726
807
|
},
|
|
@@ -739,7 +820,10 @@ var models = {
|
|
|
739
820
|
tags: ["low-cost", "reasoning", "general-purpose", "vision"],
|
|
740
821
|
lifecycle: "production",
|
|
741
822
|
capabilities: {
|
|
742
|
-
supportsImages: true
|
|
823
|
+
supportsImages: true,
|
|
824
|
+
supportsAudio: false,
|
|
825
|
+
supportsTranscription: false,
|
|
826
|
+
supportsSearch: false
|
|
743
827
|
},
|
|
744
828
|
aliases: ["gpt-5-nano"]
|
|
745
829
|
},
|
|
@@ -758,7 +842,10 @@ var models = {
|
|
|
758
842
|
tags: ["reasoning", "vision", "coding"],
|
|
759
843
|
lifecycle: "production",
|
|
760
844
|
capabilities: {
|
|
761
|
-
supportsImages: true
|
|
845
|
+
supportsImages: true,
|
|
846
|
+
supportsAudio: false,
|
|
847
|
+
supportsTranscription: false,
|
|
848
|
+
supportsSearch: false
|
|
762
849
|
},
|
|
763
850
|
aliases: ["o4-mini"]
|
|
764
851
|
},
|
|
@@ -777,7 +864,10 @@ var models = {
|
|
|
777
864
|
tags: ["reasoning", "vision", "coding"],
|
|
778
865
|
lifecycle: "production",
|
|
779
866
|
capabilities: {
|
|
780
|
-
supportsImages: true
|
|
867
|
+
supportsImages: true,
|
|
868
|
+
supportsAudio: false,
|
|
869
|
+
supportsTranscription: false,
|
|
870
|
+
supportsSearch: false
|
|
781
871
|
},
|
|
782
872
|
aliases: ["o3"]
|
|
783
873
|
},
|
|
@@ -796,7 +886,10 @@ var models = {
|
|
|
796
886
|
tags: ["recommended", "vision", "general-purpose"],
|
|
797
887
|
lifecycle: "production",
|
|
798
888
|
capabilities: {
|
|
799
|
-
supportsImages: true
|
|
889
|
+
supportsImages: true,
|
|
890
|
+
supportsAudio: false,
|
|
891
|
+
supportsTranscription: false,
|
|
892
|
+
supportsSearch: false
|
|
800
893
|
},
|
|
801
894
|
aliases: ["gpt-4.1"]
|
|
802
895
|
},
|
|
@@ -815,7 +908,10 @@ var models = {
|
|
|
815
908
|
tags: ["recommended", "vision", "general-purpose"],
|
|
816
909
|
lifecycle: "production",
|
|
817
910
|
capabilities: {
|
|
818
|
-
supportsImages: true
|
|
911
|
+
supportsImages: true,
|
|
912
|
+
supportsAudio: false,
|
|
913
|
+
supportsTranscription: false,
|
|
914
|
+
supportsSearch: false
|
|
819
915
|
},
|
|
820
916
|
aliases: ["gpt-4.1-mini"]
|
|
821
917
|
},
|
|
@@ -834,7 +930,10 @@ var models = {
|
|
|
834
930
|
tags: ["low-cost", "vision", "general-purpose"],
|
|
835
931
|
lifecycle: "production",
|
|
836
932
|
capabilities: {
|
|
837
|
-
supportsImages: true
|
|
933
|
+
supportsImages: true,
|
|
934
|
+
supportsAudio: false,
|
|
935
|
+
supportsTranscription: false,
|
|
936
|
+
supportsSearch: false
|
|
838
937
|
},
|
|
839
938
|
aliases: ["gpt-4.1-nano"]
|
|
840
939
|
},
|
|
@@ -853,7 +952,10 @@ var models = {
|
|
|
853
952
|
tags: ["reasoning", "general-purpose", "coding"],
|
|
854
953
|
lifecycle: "production",
|
|
855
954
|
capabilities: {
|
|
856
|
-
supportsImages: false
|
|
955
|
+
supportsImages: false,
|
|
956
|
+
supportsAudio: false,
|
|
957
|
+
supportsTranscription: false,
|
|
958
|
+
supportsSearch: false
|
|
857
959
|
},
|
|
858
960
|
aliases: ["o3-mini"]
|
|
859
961
|
},
|
|
@@ -872,7 +974,10 @@ var models = {
|
|
|
872
974
|
tags: ["reasoning", "vision", "general-purpose"],
|
|
873
975
|
lifecycle: "production",
|
|
874
976
|
capabilities: {
|
|
875
|
-
supportsImages: true
|
|
977
|
+
supportsImages: true,
|
|
978
|
+
supportsAudio: false,
|
|
979
|
+
supportsTranscription: false,
|
|
980
|
+
supportsSearch: false
|
|
876
981
|
}
|
|
877
982
|
},
|
|
878
983
|
"openai:o1-mini-2024-09-12": {
|
|
@@ -888,9 +993,12 @@ var models = {
|
|
|
888
993
|
costPer1MTokens: 4.4
|
|
889
994
|
},
|
|
890
995
|
tags: ["reasoning", "vision", "general-purpose"],
|
|
891
|
-
lifecycle: "
|
|
996
|
+
lifecycle: "discontinued",
|
|
892
997
|
capabilities: {
|
|
893
|
-
supportsImages: false
|
|
998
|
+
supportsImages: false,
|
|
999
|
+
supportsAudio: false,
|
|
1000
|
+
supportsTranscription: false,
|
|
1001
|
+
supportsSearch: false
|
|
894
1002
|
},
|
|
895
1003
|
aliases: ["o1-mini"]
|
|
896
1004
|
},
|
|
@@ -909,7 +1017,10 @@ var models = {
|
|
|
909
1017
|
tags: ["recommended", "vision", "low-cost", "general-purpose"],
|
|
910
1018
|
lifecycle: "production",
|
|
911
1019
|
capabilities: {
|
|
912
|
-
supportsImages: true
|
|
1020
|
+
supportsImages: true,
|
|
1021
|
+
supportsAudio: false,
|
|
1022
|
+
supportsTranscription: false,
|
|
1023
|
+
supportsSearch: false
|
|
913
1024
|
},
|
|
914
1025
|
aliases: ["gpt-4o-mini"]
|
|
915
1026
|
},
|
|
@@ -928,7 +1039,10 @@ var models = {
|
|
|
928
1039
|
tags: ["recommended", "vision", "general-purpose", "coding", "agents"],
|
|
929
1040
|
lifecycle: "production",
|
|
930
1041
|
capabilities: {
|
|
931
|
-
supportsImages: true
|
|
1042
|
+
supportsImages: true,
|
|
1043
|
+
supportsAudio: false,
|
|
1044
|
+
supportsTranscription: false,
|
|
1045
|
+
supportsSearch: false
|
|
932
1046
|
},
|
|
933
1047
|
aliases: ["gpt-4o"]
|
|
934
1048
|
},
|
|
@@ -947,7 +1061,10 @@ var models = {
|
|
|
947
1061
|
tags: ["deprecated", "vision", "general-purpose", "coding", "agents"],
|
|
948
1062
|
lifecycle: "deprecated",
|
|
949
1063
|
capabilities: {
|
|
950
|
-
supportsImages: true
|
|
1064
|
+
supportsImages: true,
|
|
1065
|
+
supportsAudio: false,
|
|
1066
|
+
supportsTranscription: false,
|
|
1067
|
+
supportsSearch: false
|
|
951
1068
|
}
|
|
952
1069
|
},
|
|
953
1070
|
"openai:gpt-4o-2024-05-13": {
|
|
@@ -965,7 +1082,10 @@ var models = {
|
|
|
965
1082
|
tags: ["deprecated", "vision", "general-purpose", "coding", "agents"],
|
|
966
1083
|
lifecycle: "deprecated",
|
|
967
1084
|
capabilities: {
|
|
968
|
-
supportsImages: true
|
|
1085
|
+
supportsImages: true,
|
|
1086
|
+
supportsAudio: false,
|
|
1087
|
+
supportsTranscription: false,
|
|
1088
|
+
supportsSearch: false
|
|
969
1089
|
}
|
|
970
1090
|
},
|
|
971
1091
|
"openai:gpt-4-turbo-2024-04-09": {
|
|
@@ -983,7 +1103,10 @@ var models = {
|
|
|
983
1103
|
tags: ["deprecated", "general-purpose", "coding", "agents"],
|
|
984
1104
|
lifecycle: "deprecated",
|
|
985
1105
|
capabilities: {
|
|
986
|
-
supportsImages: true
|
|
1106
|
+
supportsImages: true,
|
|
1107
|
+
supportsAudio: false,
|
|
1108
|
+
supportsTranscription: false,
|
|
1109
|
+
supportsSearch: false
|
|
987
1110
|
}
|
|
988
1111
|
},
|
|
989
1112
|
"openai:gpt-3.5-turbo-0125": {
|
|
@@ -1001,7 +1124,32 @@ var models = {
|
|
|
1001
1124
|
tags: ["deprecated", "general-purpose", "low-cost"],
|
|
1002
1125
|
lifecycle: "deprecated",
|
|
1003
1126
|
capabilities: {
|
|
1004
|
-
supportsImages: false
|
|
1127
|
+
supportsImages: false,
|
|
1128
|
+
supportsAudio: false,
|
|
1129
|
+
supportsTranscription: false,
|
|
1130
|
+
supportsSearch: false
|
|
1131
|
+
}
|
|
1132
|
+
},
|
|
1133
|
+
"openai:whisper-1": {
|
|
1134
|
+
id: "openai:whisper-1",
|
|
1135
|
+
name: "Whisper V2",
|
|
1136
|
+
description: "OpenAI Whisper V2 \u2014 general-purpose speech recognition model supporting 99 languages.",
|
|
1137
|
+
input: {
|
|
1138
|
+
maxTokens: 0,
|
|
1139
|
+
costPer1MTokens: 0,
|
|
1140
|
+
costPerMinute: 6e-3
|
|
1141
|
+
},
|
|
1142
|
+
output: {
|
|
1143
|
+
maxTokens: 0,
|
|
1144
|
+
costPer1MTokens: 0
|
|
1145
|
+
},
|
|
1146
|
+
tags: ["general-purpose", "speech-to-text"],
|
|
1147
|
+
lifecycle: "production",
|
|
1148
|
+
capabilities: {
|
|
1149
|
+
supportsImages: false,
|
|
1150
|
+
supportsAudio: false,
|
|
1151
|
+
supportsTranscription: true,
|
|
1152
|
+
supportsSearch: false
|
|
1005
1153
|
}
|
|
1006
1154
|
},
|
|
1007
1155
|
"anthropic:claude-opus-4-6": {
|
|
@@ -1009,7 +1157,7 @@ var models = {
|
|
|
1009
1157
|
name: "Claude Opus 4.6",
|
|
1010
1158
|
description: "Claude Opus 4.6 is the most intelligent Claude model, built for complex agents and coding workflows. It excels at long-running professional tasks, large codebases, complex refactors, and multi-step debugging with a 128K max output.",
|
|
1011
1159
|
input: {
|
|
1012
|
-
maxTokens:
|
|
1160
|
+
maxTokens: 1e6,
|
|
1013
1161
|
costPer1MTokens: 5
|
|
1014
1162
|
},
|
|
1015
1163
|
output: {
|
|
@@ -1019,7 +1167,10 @@ var models = {
|
|
|
1019
1167
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1020
1168
|
lifecycle: "production",
|
|
1021
1169
|
capabilities: {
|
|
1022
|
-
supportsImages: true
|
|
1170
|
+
supportsImages: true,
|
|
1171
|
+
supportsAudio: false,
|
|
1172
|
+
supportsTranscription: false,
|
|
1173
|
+
supportsSearch: true
|
|
1023
1174
|
}
|
|
1024
1175
|
},
|
|
1025
1176
|
"anthropic:claude-sonnet-4-6": {
|
|
@@ -1027,7 +1178,7 @@ var models = {
|
|
|
1027
1178
|
name: "Claude Sonnet 4.6",
|
|
1028
1179
|
description: "Claude Sonnet 4.6 offers the best combination of speed and intelligence in the Claude family. It features adaptive thinking for dynamic reasoning allocation, delivering fast responses for simple queries and deeper analysis for complex tasks.",
|
|
1029
1180
|
input: {
|
|
1030
|
-
maxTokens:
|
|
1181
|
+
maxTokens: 1e6,
|
|
1031
1182
|
costPer1MTokens: 3
|
|
1032
1183
|
},
|
|
1033
1184
|
output: {
|
|
@@ -1037,9 +1188,34 @@ var models = {
|
|
|
1037
1188
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1038
1189
|
lifecycle: "production",
|
|
1039
1190
|
capabilities: {
|
|
1040
|
-
supportsImages: true
|
|
1191
|
+
supportsImages: true,
|
|
1192
|
+
supportsAudio: false,
|
|
1193
|
+
supportsTranscription: false,
|
|
1194
|
+
supportsSearch: true
|
|
1041
1195
|
}
|
|
1042
1196
|
},
|
|
1197
|
+
"anthropic:claude-opus-4-5-20251101": {
|
|
1198
|
+
id: "anthropic:claude-opus-4-5-20251101",
|
|
1199
|
+
name: "Claude Opus 4.5",
|
|
1200
|
+
description: "Claude Opus 4.5 is a highly capable model with strong reasoning, coding, and agentic performance. It offers the same pricing tier as Opus 4.6 with a 200K context window.",
|
|
1201
|
+
input: {
|
|
1202
|
+
maxTokens: 2e5,
|
|
1203
|
+
costPer1MTokens: 5
|
|
1204
|
+
},
|
|
1205
|
+
output: {
|
|
1206
|
+
maxTokens: 64e3,
|
|
1207
|
+
costPer1MTokens: 25
|
|
1208
|
+
},
|
|
1209
|
+
tags: ["reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1210
|
+
lifecycle: "production",
|
|
1211
|
+
capabilities: {
|
|
1212
|
+
supportsImages: true,
|
|
1213
|
+
supportsAudio: false,
|
|
1214
|
+
supportsTranscription: false,
|
|
1215
|
+
supportsSearch: true
|
|
1216
|
+
},
|
|
1217
|
+
aliases: ["claude-opus-4-5"]
|
|
1218
|
+
},
|
|
1043
1219
|
"anthropic:claude-sonnet-4-5-20250929": {
|
|
1044
1220
|
id: "anthropic:claude-sonnet-4-5-20250929",
|
|
1045
1221
|
name: "Claude Sonnet 4.5",
|
|
@@ -1055,7 +1231,10 @@ var models = {
|
|
|
1055
1231
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1056
1232
|
lifecycle: "production",
|
|
1057
1233
|
capabilities: {
|
|
1058
|
-
supportsImages: true
|
|
1234
|
+
supportsImages: true,
|
|
1235
|
+
supportsAudio: false,
|
|
1236
|
+
supportsTranscription: false,
|
|
1237
|
+
supportsSearch: true
|
|
1059
1238
|
},
|
|
1060
1239
|
aliases: ["claude-sonnet-4-5"]
|
|
1061
1240
|
},
|
|
@@ -1074,7 +1253,10 @@ var models = {
|
|
|
1074
1253
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1075
1254
|
lifecycle: "production",
|
|
1076
1255
|
capabilities: {
|
|
1077
|
-
supportsImages: true
|
|
1256
|
+
supportsImages: true,
|
|
1257
|
+
supportsAudio: false,
|
|
1258
|
+
supportsTranscription: false,
|
|
1259
|
+
supportsSearch: false
|
|
1078
1260
|
},
|
|
1079
1261
|
aliases: ["claude-sonnet-4"]
|
|
1080
1262
|
},
|
|
@@ -1093,7 +1275,10 @@ var models = {
|
|
|
1093
1275
|
tags: ["deprecated", "vision", "reasoning", "general-purpose", "agents", "coding"],
|
|
1094
1276
|
lifecycle: "deprecated",
|
|
1095
1277
|
capabilities: {
|
|
1096
|
-
supportsImages: true
|
|
1278
|
+
supportsImages: true,
|
|
1279
|
+
supportsAudio: false,
|
|
1280
|
+
supportsTranscription: false,
|
|
1281
|
+
supportsSearch: false
|
|
1097
1282
|
},
|
|
1098
1283
|
aliases: ["claude-sonnet-4-reasoning"]
|
|
1099
1284
|
},
|
|
@@ -1112,7 +1297,10 @@ var models = {
|
|
|
1112
1297
|
tags: ["recommended", "agents", "vision", "general-purpose", "coding"],
|
|
1113
1298
|
lifecycle: "production",
|
|
1114
1299
|
capabilities: {
|
|
1115
|
-
supportsImages: true
|
|
1300
|
+
supportsImages: true,
|
|
1301
|
+
supportsAudio: false,
|
|
1302
|
+
supportsTranscription: false,
|
|
1303
|
+
supportsSearch: false
|
|
1116
1304
|
},
|
|
1117
1305
|
aliases: ["claude-haiku-4-5"]
|
|
1118
1306
|
},
|
|
@@ -1133,9 +1321,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1133
1321
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1134
1322
|
lifecycle: "production",
|
|
1135
1323
|
capabilities: {
|
|
1136
|
-
supportsImages: true
|
|
1324
|
+
supportsImages: true,
|
|
1325
|
+
supportsAudio: false,
|
|
1326
|
+
supportsTranscription: false,
|
|
1327
|
+
supportsSearch: false
|
|
1137
1328
|
},
|
|
1138
|
-
aliases: ["claude-haiku-4-5-reasoning"]
|
|
1329
|
+
aliases: ["claude-haiku-4-5-reasoning", "claude-haiku-4-5-20251001"]
|
|
1139
1330
|
},
|
|
1140
1331
|
"anthropic:claude-3-7-sonnet-20250219": {
|
|
1141
1332
|
id: "anthropic:claude-3-7-sonnet-20250219",
|
|
@@ -1152,7 +1343,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1152
1343
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1153
1344
|
lifecycle: "discontinued",
|
|
1154
1345
|
capabilities: {
|
|
1155
|
-
supportsImages: true
|
|
1346
|
+
supportsImages: true,
|
|
1347
|
+
supportsAudio: false,
|
|
1348
|
+
supportsTranscription: false,
|
|
1349
|
+
supportsSearch: false
|
|
1156
1350
|
}
|
|
1157
1351
|
},
|
|
1158
1352
|
"anthropic:claude-3-7-sonnet-reasoning-20250219": {
|
|
@@ -1170,7 +1364,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1170
1364
|
tags: ["deprecated", "vision", "reasoning", "general-purpose", "agents", "coding"],
|
|
1171
1365
|
lifecycle: "discontinued",
|
|
1172
1366
|
capabilities: {
|
|
1173
|
-
supportsImages: true
|
|
1367
|
+
supportsImages: true,
|
|
1368
|
+
supportsAudio: false,
|
|
1369
|
+
supportsTranscription: false,
|
|
1370
|
+
supportsSearch: false
|
|
1174
1371
|
}
|
|
1175
1372
|
},
|
|
1176
1373
|
"anthropic:claude-3-5-haiku-20241022": {
|
|
@@ -1188,7 +1385,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1188
1385
|
tags: ["general-purpose", "low-cost"],
|
|
1189
1386
|
lifecycle: "discontinued",
|
|
1190
1387
|
capabilities: {
|
|
1191
|
-
supportsImages: true
|
|
1388
|
+
supportsImages: true,
|
|
1389
|
+
supportsAudio: false,
|
|
1390
|
+
supportsTranscription: false,
|
|
1391
|
+
supportsSearch: false
|
|
1192
1392
|
}
|
|
1193
1393
|
},
|
|
1194
1394
|
"anthropic:claude-3-5-sonnet-20241022": {
|
|
@@ -1206,7 +1406,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1206
1406
|
tags: ["vision", "general-purpose", "agents", "coding", "storytelling"],
|
|
1207
1407
|
lifecycle: "discontinued",
|
|
1208
1408
|
capabilities: {
|
|
1209
|
-
supportsImages: true
|
|
1409
|
+
supportsImages: true,
|
|
1410
|
+
supportsAudio: false,
|
|
1411
|
+
supportsTranscription: false,
|
|
1412
|
+
supportsSearch: false
|
|
1210
1413
|
}
|
|
1211
1414
|
},
|
|
1212
1415
|
"anthropic:claude-3-5-sonnet-20240620": {
|
|
@@ -1224,7 +1427,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1224
1427
|
tags: ["vision", "general-purpose", "agents", "coding", "storytelling"],
|
|
1225
1428
|
lifecycle: "discontinued",
|
|
1226
1429
|
capabilities: {
|
|
1227
|
-
supportsImages: true
|
|
1430
|
+
supportsImages: true,
|
|
1431
|
+
supportsAudio: false,
|
|
1432
|
+
supportsTranscription: false,
|
|
1433
|
+
supportsSearch: false
|
|
1228
1434
|
}
|
|
1229
1435
|
},
|
|
1230
1436
|
"anthropic:claude-3-haiku-20240307": {
|
|
@@ -1242,13 +1448,16 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1242
1448
|
tags: ["deprecated", "low-cost", "general-purpose"],
|
|
1243
1449
|
lifecycle: "deprecated",
|
|
1244
1450
|
capabilities: {
|
|
1245
|
-
supportsImages: true
|
|
1451
|
+
supportsImages: true,
|
|
1452
|
+
supportsAudio: false,
|
|
1453
|
+
supportsTranscription: false,
|
|
1454
|
+
supportsSearch: false
|
|
1246
1455
|
}
|
|
1247
1456
|
},
|
|
1248
|
-
"google-ai:gemini-3-pro": {
|
|
1249
|
-
id: "google-ai:gemini-3-pro",
|
|
1250
|
-
name: "Gemini 3 Pro",
|
|
1251
|
-
description: "
|
|
1457
|
+
"google-ai:gemini-3.1-pro": {
|
|
1458
|
+
id: "google-ai:gemini-3.1-pro",
|
|
1459
|
+
name: "Gemini 3.1 Pro",
|
|
1460
|
+
description: "Google's most powerful agentic and coding model, delivering state-of-the-art reasoning with rich multimodal understanding. Successor to Gemini 3 Pro (shut down March 9, 2026).",
|
|
1252
1461
|
input: {
|
|
1253
1462
|
maxTokens: 1048576,
|
|
1254
1463
|
costPer1MTokens: 2
|
|
@@ -1257,10 +1466,13 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1257
1466
|
maxTokens: 65536,
|
|
1258
1467
|
costPer1MTokens: 12
|
|
1259
1468
|
},
|
|
1260
|
-
tags: ["reasoning", "agents", "general-purpose", "vision"],
|
|
1469
|
+
tags: ["reasoning", "agents", "general-purpose", "vision", "coding"],
|
|
1261
1470
|
lifecycle: "preview",
|
|
1262
1471
|
capabilities: {
|
|
1263
|
-
supportsImages: true
|
|
1472
|
+
supportsImages: true,
|
|
1473
|
+
supportsAudio: true,
|
|
1474
|
+
supportsTranscription: false,
|
|
1475
|
+
supportsSearch: true
|
|
1264
1476
|
},
|
|
1265
1477
|
aliases: ["gemini-3.1-pro-preview"]
|
|
1266
1478
|
},
|
|
@@ -1279,10 +1491,56 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1279
1491
|
tags: ["reasoning", "agents", "general-purpose", "vision"],
|
|
1280
1492
|
lifecycle: "preview",
|
|
1281
1493
|
capabilities: {
|
|
1282
|
-
supportsImages: true
|
|
1494
|
+
supportsImages: true,
|
|
1495
|
+
supportsAudio: true,
|
|
1496
|
+
supportsTranscription: false,
|
|
1497
|
+
supportsSearch: true
|
|
1283
1498
|
},
|
|
1284
1499
|
aliases: ["gemini-3-flash-preview"]
|
|
1285
1500
|
},
|
|
1501
|
+
"google-ai:gemini-3.1-flash-lite": {
|
|
1502
|
+
id: "google-ai:gemini-3.1-flash-lite",
|
|
1503
|
+
name: "Gemini 3.1 Flash-Lite",
|
|
1504
|
+
description: "Google's most cost-effective AI model for high-volume, low-latency tasks. Offers strong performance at a fraction of the cost of larger models.",
|
|
1505
|
+
input: {
|
|
1506
|
+
maxTokens: 1048576,
|
|
1507
|
+
costPer1MTokens: 0.25
|
|
1508
|
+
},
|
|
1509
|
+
output: {
|
|
1510
|
+
maxTokens: 65536,
|
|
1511
|
+
costPer1MTokens: 1.5
|
|
1512
|
+
},
|
|
1513
|
+
tags: ["low-cost", "general-purpose", "vision"],
|
|
1514
|
+
lifecycle: "preview",
|
|
1515
|
+
capabilities: {
|
|
1516
|
+
supportsImages: true,
|
|
1517
|
+
supportsAudio: true,
|
|
1518
|
+
supportsTranscription: false,
|
|
1519
|
+
supportsSearch: true
|
|
1520
|
+
},
|
|
1521
|
+
aliases: ["gemini-3.1-flash-lite-preview"]
|
|
1522
|
+
},
|
|
1523
|
+
"google-ai:gemini-2.5-pro": {
|
|
1524
|
+
id: "google-ai:gemini-2.5-pro",
|
|
1525
|
+
name: "Gemini 2.5 Pro",
|
|
1526
|
+
description: `Google's most advanced stable AI model designed for complex reasoning, coding, mathematics, and scientific tasks. Features "thinking" capabilities for superior human-preference alignment and problem-solving.`,
|
|
1527
|
+
input: {
|
|
1528
|
+
maxTokens: 2e5,
|
|
1529
|
+
costPer1MTokens: 1.25
|
|
1530
|
+
},
|
|
1531
|
+
output: {
|
|
1532
|
+
maxTokens: 65536,
|
|
1533
|
+
costPer1MTokens: 10
|
|
1534
|
+
},
|
|
1535
|
+
tags: ["recommended", "reasoning", "agents", "general-purpose", "vision", "coding"],
|
|
1536
|
+
lifecycle: "production",
|
|
1537
|
+
capabilities: {
|
|
1538
|
+
supportsImages: true,
|
|
1539
|
+
supportsAudio: true,
|
|
1540
|
+
supportsTranscription: false,
|
|
1541
|
+
supportsSearch: true
|
|
1542
|
+
}
|
|
1543
|
+
},
|
|
1286
1544
|
"google-ai:gemini-2.5-flash": {
|
|
1287
1545
|
id: "google-ai:gemini-2.5-flash",
|
|
1288
1546
|
name: "Gemini 2.5 Flash",
|
|
@@ -1298,25 +1556,31 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1298
1556
|
tags: ["recommended", "reasoning", "agents", "general-purpose", "vision"],
|
|
1299
1557
|
lifecycle: "production",
|
|
1300
1558
|
capabilities: {
|
|
1301
|
-
supportsImages: true
|
|
1559
|
+
supportsImages: true,
|
|
1560
|
+
supportsAudio: true,
|
|
1561
|
+
supportsTranscription: false,
|
|
1562
|
+
supportsSearch: true
|
|
1302
1563
|
}
|
|
1303
1564
|
},
|
|
1304
|
-
"google-ai:gemini-2.5-
|
|
1305
|
-
id: "google-ai:gemini-2.5-
|
|
1306
|
-
name: "Gemini 2.5
|
|
1307
|
-
description:
|
|
1565
|
+
"google-ai:gemini-2.5-flash-lite": {
|
|
1566
|
+
id: "google-ai:gemini-2.5-flash-lite",
|
|
1567
|
+
name: "Gemini 2.5 Flash-Lite",
|
|
1568
|
+
description: "Lightweight, cost-efficient Gemini model optimized for high-volume, low-latency tasks. Successor to Gemini 2.0 Flash with improved capabilities.",
|
|
1308
1569
|
input: {
|
|
1309
|
-
maxTokens:
|
|
1310
|
-
costPer1MTokens: 1
|
|
1570
|
+
maxTokens: 1048576,
|
|
1571
|
+
costPer1MTokens: 0.1
|
|
1311
1572
|
},
|
|
1312
1573
|
output: {
|
|
1313
1574
|
maxTokens: 65536,
|
|
1314
|
-
costPer1MTokens:
|
|
1575
|
+
costPer1MTokens: 0.4
|
|
1315
1576
|
},
|
|
1316
|
-
tags: ["recommended", "
|
|
1577
|
+
tags: ["recommended", "low-cost", "general-purpose", "vision"],
|
|
1317
1578
|
lifecycle: "production",
|
|
1318
1579
|
capabilities: {
|
|
1319
|
-
supportsImages: true
|
|
1580
|
+
supportsImages: true,
|
|
1581
|
+
supportsAudio: true,
|
|
1582
|
+
supportsTranscription: false,
|
|
1583
|
+
supportsSearch: true
|
|
1320
1584
|
}
|
|
1321
1585
|
},
|
|
1322
1586
|
"google-ai:gemini-2.0-flash": {
|
|
@@ -1332,12 +1596,37 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1332
1596
|
costPer1MTokens: 0.4
|
|
1333
1597
|
},
|
|
1334
1598
|
tags: ["low-cost", "general-purpose", "vision"],
|
|
1335
|
-
lifecycle: "
|
|
1599
|
+
lifecycle: "deprecated",
|
|
1336
1600
|
capabilities: {
|
|
1337
|
-
supportsImages: true
|
|
1601
|
+
supportsImages: true,
|
|
1602
|
+
supportsAudio: true,
|
|
1603
|
+
supportsTranscription: false,
|
|
1604
|
+
supportsSearch: false
|
|
1338
1605
|
},
|
|
1339
1606
|
aliases: ["models/gemini-2.0-flash"]
|
|
1340
1607
|
},
|
|
1608
|
+
"google-ai:gemini-3-pro": {
|
|
1609
|
+
id: "google-ai:gemini-3-pro",
|
|
1610
|
+
name: "Gemini 3 Pro (Shut Down)",
|
|
1611
|
+
description: "Gemini 3 Pro Preview was shut down on March 9, 2026. Use Gemini 3.1 Pro instead.",
|
|
1612
|
+
input: {
|
|
1613
|
+
maxTokens: 1048576,
|
|
1614
|
+
costPer1MTokens: 2
|
|
1615
|
+
},
|
|
1616
|
+
output: {
|
|
1617
|
+
maxTokens: 65536,
|
|
1618
|
+
costPer1MTokens: 12
|
|
1619
|
+
},
|
|
1620
|
+
tags: ["reasoning", "agents", "general-purpose", "vision"],
|
|
1621
|
+
lifecycle: "discontinued",
|
|
1622
|
+
capabilities: {
|
|
1623
|
+
supportsImages: true,
|
|
1624
|
+
supportsAudio: true,
|
|
1625
|
+
supportsTranscription: false,
|
|
1626
|
+
supportsSearch: true
|
|
1627
|
+
},
|
|
1628
|
+
aliases: ["gemini-3-pro-preview"]
|
|
1629
|
+
},
|
|
1341
1630
|
"cerebras:gpt-oss-120b": {
|
|
1342
1631
|
id: "cerebras:gpt-oss-120b",
|
|
1343
1632
|
name: "GPT-OSS 120B (Preview)",
|
|
@@ -1353,7 +1642,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1353
1642
|
tags: ["preview", "general-purpose", "reasoning"],
|
|
1354
1643
|
lifecycle: "production",
|
|
1355
1644
|
capabilities: {
|
|
1356
|
-
supportsImages: false
|
|
1645
|
+
supportsImages: false,
|
|
1646
|
+
supportsAudio: false,
|
|
1647
|
+
supportsTranscription: false,
|
|
1648
|
+
supportsSearch: false
|
|
1357
1649
|
}
|
|
1358
1650
|
},
|
|
1359
1651
|
"cerebras:qwen-3-32b": {
|
|
@@ -1369,9 +1661,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1369
1661
|
costPer1MTokens: 0.8
|
|
1370
1662
|
},
|
|
1371
1663
|
tags: ["general-purpose", "reasoning"],
|
|
1372
|
-
lifecycle: "
|
|
1664
|
+
lifecycle: "discontinued",
|
|
1373
1665
|
capabilities: {
|
|
1374
|
-
supportsImages: false
|
|
1666
|
+
supportsImages: false,
|
|
1667
|
+
supportsAudio: false,
|
|
1668
|
+
supportsTranscription: false,
|
|
1669
|
+
supportsSearch: false
|
|
1375
1670
|
}
|
|
1376
1671
|
},
|
|
1377
1672
|
"cerebras:llama-4-scout-17b-16e-instruct": {
|
|
@@ -1387,9 +1682,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1387
1682
|
costPer1MTokens: 0.85
|
|
1388
1683
|
},
|
|
1389
1684
|
tags: ["general-purpose", "vision"],
|
|
1390
|
-
lifecycle: "
|
|
1685
|
+
lifecycle: "discontinued",
|
|
1391
1686
|
capabilities: {
|
|
1392
|
-
supportsImages: true
|
|
1687
|
+
supportsImages: true,
|
|
1688
|
+
supportsAudio: false,
|
|
1689
|
+
supportsTranscription: false,
|
|
1690
|
+
supportsSearch: false
|
|
1393
1691
|
}
|
|
1394
1692
|
},
|
|
1395
1693
|
"cerebras:llama3.1-8b": {
|
|
@@ -1407,7 +1705,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1407
1705
|
tags: ["low-cost", "general-purpose"],
|
|
1408
1706
|
lifecycle: "production",
|
|
1409
1707
|
capabilities: {
|
|
1410
|
-
supportsImages: false
|
|
1708
|
+
supportsImages: false,
|
|
1709
|
+
supportsAudio: false,
|
|
1710
|
+
supportsTranscription: false,
|
|
1711
|
+
supportsSearch: false
|
|
1411
1712
|
}
|
|
1412
1713
|
},
|
|
1413
1714
|
"cerebras:llama3.3-70b": {
|
|
@@ -1423,9 +1724,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1423
1724
|
costPer1MTokens: 1.2
|
|
1424
1725
|
},
|
|
1425
1726
|
tags: ["general-purpose"],
|
|
1426
|
-
lifecycle: "
|
|
1727
|
+
lifecycle: "discontinued",
|
|
1427
1728
|
capabilities: {
|
|
1428
|
-
supportsImages: false
|
|
1729
|
+
supportsImages: false,
|
|
1730
|
+
supportsAudio: false,
|
|
1731
|
+
supportsTranscription: false,
|
|
1732
|
+
supportsSearch: false
|
|
1429
1733
|
}
|
|
1430
1734
|
},
|
|
1431
1735
|
"groq:gpt-oss-20b": {
|
|
@@ -1434,16 +1738,19 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1434
1738
|
description: "gpt-oss-20b is a compact, open-weight language model optimized for low-latency. It shares the same training foundation and capabilities as the GPT-OSS 120B model, with faster responses and lower cost.",
|
|
1435
1739
|
input: {
|
|
1436
1740
|
maxTokens: 131e3,
|
|
1437
|
-
costPer1MTokens: 0.
|
|
1741
|
+
costPer1MTokens: 0.075
|
|
1438
1742
|
},
|
|
1439
1743
|
output: {
|
|
1440
1744
|
maxTokens: 32e3,
|
|
1441
|
-
costPer1MTokens: 0.
|
|
1745
|
+
costPer1MTokens: 0.3
|
|
1442
1746
|
},
|
|
1443
1747
|
tags: ["preview", "general-purpose", "reasoning", "low-cost"],
|
|
1444
1748
|
lifecycle: "production",
|
|
1445
1749
|
capabilities: {
|
|
1446
|
-
supportsImages: false
|
|
1750
|
+
supportsImages: false,
|
|
1751
|
+
supportsAudio: false,
|
|
1752
|
+
supportsTranscription: false,
|
|
1753
|
+
supportsSearch: false
|
|
1447
1754
|
},
|
|
1448
1755
|
aliases: ["openai/gpt-oss-20b"]
|
|
1449
1756
|
},
|
|
@@ -1462,7 +1769,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1462
1769
|
tags: ["preview", "general-purpose", "reasoning"],
|
|
1463
1770
|
lifecycle: "production",
|
|
1464
1771
|
capabilities: {
|
|
1465
|
-
supportsImages: false
|
|
1772
|
+
supportsImages: false,
|
|
1773
|
+
supportsAudio: false,
|
|
1774
|
+
supportsTranscription: false,
|
|
1775
|
+
supportsSearch: false
|
|
1466
1776
|
},
|
|
1467
1777
|
aliases: ["openai/gpt-oss-120b"]
|
|
1468
1778
|
},
|
|
@@ -1481,7 +1791,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1481
1791
|
tags: ["general-purpose", "reasoning", "preview"],
|
|
1482
1792
|
lifecycle: "discontinued",
|
|
1483
1793
|
capabilities: {
|
|
1484
|
-
supportsImages: false
|
|
1794
|
+
supportsImages: false,
|
|
1795
|
+
supportsAudio: false,
|
|
1796
|
+
supportsTranscription: false,
|
|
1797
|
+
supportsSearch: false
|
|
1485
1798
|
}
|
|
1486
1799
|
},
|
|
1487
1800
|
"groq:llama-3.3-70b-versatile": {
|
|
@@ -1499,7 +1812,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1499
1812
|
tags: ["recommended", "general-purpose", "coding"],
|
|
1500
1813
|
lifecycle: "production",
|
|
1501
1814
|
capabilities: {
|
|
1502
|
-
supportsImages: false
|
|
1815
|
+
supportsImages: false,
|
|
1816
|
+
supportsAudio: false,
|
|
1817
|
+
supportsTranscription: false,
|
|
1818
|
+
supportsSearch: false
|
|
1503
1819
|
}
|
|
1504
1820
|
},
|
|
1505
1821
|
"groq:llama-3.2-1b-preview": {
|
|
@@ -1517,7 +1833,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1517
1833
|
tags: ["low-cost", "deprecated"],
|
|
1518
1834
|
lifecycle: "discontinued",
|
|
1519
1835
|
capabilities: {
|
|
1520
|
-
supportsImages: false
|
|
1836
|
+
supportsImages: false,
|
|
1837
|
+
supportsAudio: false,
|
|
1838
|
+
supportsTranscription: false,
|
|
1839
|
+
supportsSearch: false
|
|
1521
1840
|
}
|
|
1522
1841
|
},
|
|
1523
1842
|
"groq:llama-3.2-3b-preview": {
|
|
@@ -1535,7 +1854,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1535
1854
|
tags: ["low-cost", "general-purpose", "deprecated"],
|
|
1536
1855
|
lifecycle: "discontinued",
|
|
1537
1856
|
capabilities: {
|
|
1538
|
-
supportsImages: false
|
|
1857
|
+
supportsImages: false,
|
|
1858
|
+
supportsAudio: false,
|
|
1859
|
+
supportsTranscription: false,
|
|
1860
|
+
supportsSearch: false
|
|
1539
1861
|
}
|
|
1540
1862
|
},
|
|
1541
1863
|
"groq:llama-3.2-11b-vision-preview": {
|
|
@@ -1553,7 +1875,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1553
1875
|
tags: ["low-cost", "vision", "general-purpose", "deprecated"],
|
|
1554
1876
|
lifecycle: "discontinued",
|
|
1555
1877
|
capabilities: {
|
|
1556
|
-
supportsImages: true
|
|
1878
|
+
supportsImages: true,
|
|
1879
|
+
supportsAudio: false,
|
|
1880
|
+
supportsTranscription: false,
|
|
1881
|
+
supportsSearch: false
|
|
1557
1882
|
}
|
|
1558
1883
|
},
|
|
1559
1884
|
"groq:llama-3.2-90b-vision-preview": {
|
|
@@ -1571,7 +1896,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1571
1896
|
tags: ["vision", "general-purpose", "deprecated"],
|
|
1572
1897
|
lifecycle: "discontinued",
|
|
1573
1898
|
capabilities: {
|
|
1574
|
-
supportsImages: true
|
|
1899
|
+
supportsImages: true,
|
|
1900
|
+
supportsAudio: false,
|
|
1901
|
+
supportsTranscription: false,
|
|
1902
|
+
supportsSearch: false
|
|
1575
1903
|
}
|
|
1576
1904
|
},
|
|
1577
1905
|
"groq:llama-3.1-8b-instant": {
|
|
@@ -1589,7 +1917,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1589
1917
|
tags: ["low-cost", "general-purpose"],
|
|
1590
1918
|
lifecycle: "production",
|
|
1591
1919
|
capabilities: {
|
|
1592
|
-
supportsImages: false
|
|
1920
|
+
supportsImages: false,
|
|
1921
|
+
supportsAudio: false,
|
|
1922
|
+
supportsTranscription: false,
|
|
1923
|
+
supportsSearch: false
|
|
1593
1924
|
}
|
|
1594
1925
|
},
|
|
1595
1926
|
"groq:llama3-8b-8192": {
|
|
@@ -1607,7 +1938,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1607
1938
|
tags: ["low-cost", "general-purpose", "deprecated"],
|
|
1608
1939
|
lifecycle: "discontinued",
|
|
1609
1940
|
capabilities: {
|
|
1610
|
-
supportsImages: false
|
|
1941
|
+
supportsImages: false,
|
|
1942
|
+
supportsAudio: false,
|
|
1943
|
+
supportsTranscription: false,
|
|
1944
|
+
supportsSearch: false
|
|
1611
1945
|
}
|
|
1612
1946
|
},
|
|
1613
1947
|
"groq:llama3-70b-8192": {
|
|
@@ -1625,7 +1959,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1625
1959
|
tags: ["general-purpose", "deprecated"],
|
|
1626
1960
|
lifecycle: "discontinued",
|
|
1627
1961
|
capabilities: {
|
|
1628
|
-
supportsImages: false
|
|
1962
|
+
supportsImages: false,
|
|
1963
|
+
supportsAudio: false,
|
|
1964
|
+
supportsTranscription: false,
|
|
1965
|
+
supportsSearch: false
|
|
1629
1966
|
}
|
|
1630
1967
|
},
|
|
1631
1968
|
"groq:gemma2-9b-it": {
|
|
@@ -1643,7 +1980,160 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1643
1980
|
tags: ["low-cost", "general-purpose"],
|
|
1644
1981
|
lifecycle: "discontinued",
|
|
1645
1982
|
capabilities: {
|
|
1646
|
-
supportsImages: false
|
|
1983
|
+
supportsImages: false,
|
|
1984
|
+
supportsAudio: false,
|
|
1985
|
+
supportsTranscription: false,
|
|
1986
|
+
supportsSearch: false
|
|
1987
|
+
}
|
|
1988
|
+
},
|
|
1989
|
+
"groq:whisper-large-v3": {
|
|
1990
|
+
id: "groq:whisper-large-v3",
|
|
1991
|
+
name: "Whisper V3",
|
|
1992
|
+
description: "Whisper Large V3 on Groq \u2014 fast, accurate multilingual speech recognition.",
|
|
1993
|
+
input: {
|
|
1994
|
+
maxTokens: 0,
|
|
1995
|
+
costPer1MTokens: 0,
|
|
1996
|
+
costPerMinute: 185e-5
|
|
1997
|
+
},
|
|
1998
|
+
output: {
|
|
1999
|
+
maxTokens: 0,
|
|
2000
|
+
costPer1MTokens: 0
|
|
2001
|
+
},
|
|
2002
|
+
tags: ["general-purpose", "speech-to-text"],
|
|
2003
|
+
lifecycle: "production",
|
|
2004
|
+
capabilities: {
|
|
2005
|
+
supportsImages: false,
|
|
2006
|
+
supportsAudio: false,
|
|
2007
|
+
supportsTranscription: true,
|
|
2008
|
+
supportsSearch: false
|
|
2009
|
+
}
|
|
2010
|
+
},
|
|
2011
|
+
"groq:whisper-large-v3-turbo": {
|
|
2012
|
+
id: "groq:whisper-large-v3-turbo",
|
|
2013
|
+
name: "Whisper V3 Turbo",
|
|
2014
|
+
description: "Whisper Large V3 Turbo on Groq \u2014 optimized for speed with near-identical accuracy to V3.",
|
|
2015
|
+
input: {
|
|
2016
|
+
maxTokens: 0,
|
|
2017
|
+
costPer1MTokens: 0,
|
|
2018
|
+
costPerMinute: 667e-6
|
|
2019
|
+
},
|
|
2020
|
+
output: {
|
|
2021
|
+
maxTokens: 0,
|
|
2022
|
+
costPer1MTokens: 0
|
|
2023
|
+
},
|
|
2024
|
+
tags: ["low-cost", "general-purpose", "speech-to-text"],
|
|
2025
|
+
lifecycle: "production",
|
|
2026
|
+
capabilities: {
|
|
2027
|
+
supportsImages: false,
|
|
2028
|
+
supportsAudio: false,
|
|
2029
|
+
supportsTranscription: true,
|
|
2030
|
+
supportsSearch: false
|
|
2031
|
+
}
|
|
2032
|
+
},
|
|
2033
|
+
"groq:distil-whisper-large-v3-en": {
|
|
2034
|
+
id: "groq:distil-whisper-large-v3-en",
|
|
2035
|
+
name: "Distil Whisper V3 (English)",
|
|
2036
|
+
description: "Distilled Whisper Large V3 on Groq \u2014 decommissioned, replaced by whisper-large-v3-turbo.",
|
|
2037
|
+
input: {
|
|
2038
|
+
maxTokens: 0,
|
|
2039
|
+
costPer1MTokens: 0,
|
|
2040
|
+
costPerMinute: 333e-6
|
|
2041
|
+
},
|
|
2042
|
+
output: {
|
|
2043
|
+
maxTokens: 0,
|
|
2044
|
+
costPer1MTokens: 0
|
|
2045
|
+
},
|
|
2046
|
+
tags: ["low-cost", "speech-to-text"],
|
|
2047
|
+
lifecycle: "discontinued",
|
|
2048
|
+
capabilities: {
|
|
2049
|
+
supportsImages: false,
|
|
2050
|
+
supportsAudio: false,
|
|
2051
|
+
supportsTranscription: true,
|
|
2052
|
+
supportsSearch: false
|
|
2053
|
+
}
|
|
2054
|
+
},
|
|
2055
|
+
"xai:grok-4-1-fast-reasoning": {
|
|
2056
|
+
id: "xai:grok-4-1-fast-reasoning",
|
|
2057
|
+
name: "Grok 4.1 Fast (Reasoning)",
|
|
2058
|
+
description: "Latest fast Grok model with reasoning capabilities and a massive 2M context window. Extremely cost-effective for a frontier-class model.",
|
|
2059
|
+
input: {
|
|
2060
|
+
maxTokens: 2e6,
|
|
2061
|
+
costPer1MTokens: 0.2
|
|
2062
|
+
},
|
|
2063
|
+
output: {
|
|
2064
|
+
maxTokens: 128e3,
|
|
2065
|
+
costPer1MTokens: 0.5
|
|
2066
|
+
},
|
|
2067
|
+
tags: ["recommended", "reasoning", "general-purpose", "vision", "low-cost"],
|
|
2068
|
+
lifecycle: "production",
|
|
2069
|
+
capabilities: {
|
|
2070
|
+
supportsImages: true,
|
|
2071
|
+
supportsAudio: false,
|
|
2072
|
+
supportsTranscription: false,
|
|
2073
|
+
supportsSearch: false
|
|
2074
|
+
}
|
|
2075
|
+
},
|
|
2076
|
+
"xai:grok-4-1-fast-non-reasoning": {
|
|
2077
|
+
id: "xai:grok-4-1-fast-non-reasoning",
|
|
2078
|
+
name: "Grok 4.1 Fast (Non-Reasoning)",
|
|
2079
|
+
description: "Latest fast Grok model for non-reasoning tasks with a massive 2M context window. Extremely cost-effective for a frontier-class model.",
|
|
2080
|
+
input: {
|
|
2081
|
+
maxTokens: 2e6,
|
|
2082
|
+
costPer1MTokens: 0.2
|
|
2083
|
+
},
|
|
2084
|
+
output: {
|
|
2085
|
+
maxTokens: 128e3,
|
|
2086
|
+
costPer1MTokens: 0.5
|
|
2087
|
+
},
|
|
2088
|
+
tags: ["recommended", "general-purpose", "vision", "low-cost"],
|
|
2089
|
+
lifecycle: "production",
|
|
2090
|
+
capabilities: {
|
|
2091
|
+
supportsImages: true,
|
|
2092
|
+
supportsAudio: false,
|
|
2093
|
+
supportsTranscription: false,
|
|
2094
|
+
supportsSearch: false
|
|
2095
|
+
}
|
|
2096
|
+
},
|
|
2097
|
+
"xai:grok-4.20-0309-reasoning": {
|
|
2098
|
+
id: "xai:grok-4.20-0309-reasoning",
|
|
2099
|
+
name: "Grok 4.20 (Reasoning)",
|
|
2100
|
+
description: "xAI flagship model with deep reasoning capabilities and 2M context window.",
|
|
2101
|
+
input: {
|
|
2102
|
+
maxTokens: 2e6,
|
|
2103
|
+
costPer1MTokens: 2
|
|
2104
|
+
},
|
|
2105
|
+
output: {
|
|
2106
|
+
maxTokens: 128e3,
|
|
2107
|
+
costPer1MTokens: 6
|
|
2108
|
+
},
|
|
2109
|
+
tags: ["reasoning", "general-purpose", "vision"],
|
|
2110
|
+
lifecycle: "production",
|
|
2111
|
+
capabilities: {
|
|
2112
|
+
supportsImages: true,
|
|
2113
|
+
supportsAudio: false,
|
|
2114
|
+
supportsTranscription: false,
|
|
2115
|
+
supportsSearch: false
|
|
2116
|
+
}
|
|
2117
|
+
},
|
|
2118
|
+
"xai:grok-4.20-0309-non-reasoning": {
|
|
2119
|
+
id: "xai:grok-4.20-0309-non-reasoning",
|
|
2120
|
+
name: "Grok 4.20 (Non-Reasoning)",
|
|
2121
|
+
description: "xAI flagship model for non-reasoning tasks with 2M context window.",
|
|
2122
|
+
input: {
|
|
2123
|
+
maxTokens: 2e6,
|
|
2124
|
+
costPer1MTokens: 2
|
|
2125
|
+
},
|
|
2126
|
+
output: {
|
|
2127
|
+
maxTokens: 128e3,
|
|
2128
|
+
costPer1MTokens: 6
|
|
2129
|
+
},
|
|
2130
|
+
tags: ["general-purpose", "vision"],
|
|
2131
|
+
lifecycle: "production",
|
|
2132
|
+
capabilities: {
|
|
2133
|
+
supportsImages: true,
|
|
2134
|
+
supportsAudio: false,
|
|
2135
|
+
supportsTranscription: false,
|
|
2136
|
+
supportsSearch: false
|
|
1647
2137
|
}
|
|
1648
2138
|
},
|
|
1649
2139
|
"xai:grok-code-fast-1": {
|
|
@@ -1661,7 +2151,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1661
2151
|
tags: ["coding", "general-purpose", "low-cost"],
|
|
1662
2152
|
lifecycle: "production",
|
|
1663
2153
|
capabilities: {
|
|
1664
|
-
supportsImages: false
|
|
2154
|
+
supportsImages: false,
|
|
2155
|
+
supportsAudio: false,
|
|
2156
|
+
supportsTranscription: false,
|
|
2157
|
+
supportsSearch: false
|
|
1665
2158
|
}
|
|
1666
2159
|
},
|
|
1667
2160
|
"xai:grok-4-fast-reasoning": {
|
|
@@ -1676,10 +2169,13 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1676
2169
|
maxTokens: 128e3,
|
|
1677
2170
|
costPer1MTokens: 0.5
|
|
1678
2171
|
},
|
|
1679
|
-
tags: ["reasoning", "
|
|
2172
|
+
tags: ["reasoning", "general-purpose"],
|
|
1680
2173
|
lifecycle: "production",
|
|
1681
2174
|
capabilities: {
|
|
1682
|
-
supportsImages: true
|
|
2175
|
+
supportsImages: true,
|
|
2176
|
+
supportsAudio: false,
|
|
2177
|
+
supportsTranscription: false,
|
|
2178
|
+
supportsSearch: false
|
|
1683
2179
|
}
|
|
1684
2180
|
},
|
|
1685
2181
|
"xai:grok-4-fast-non-reasoning": {
|
|
@@ -1694,10 +2190,13 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1694
2190
|
maxTokens: 128e3,
|
|
1695
2191
|
costPer1MTokens: 0.5
|
|
1696
2192
|
},
|
|
1697
|
-
tags: ["low-cost", "
|
|
2193
|
+
tags: ["low-cost", "general-purpose"],
|
|
1698
2194
|
lifecycle: "production",
|
|
1699
2195
|
capabilities: {
|
|
1700
|
-
supportsImages: true
|
|
2196
|
+
supportsImages: true,
|
|
2197
|
+
supportsAudio: false,
|
|
2198
|
+
supportsTranscription: false,
|
|
2199
|
+
supportsSearch: false
|
|
1701
2200
|
}
|
|
1702
2201
|
},
|
|
1703
2202
|
"xai:grok-4-0709": {
|
|
@@ -1715,7 +2214,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1715
2214
|
tags: ["reasoning", "general-purpose"],
|
|
1716
2215
|
lifecycle: "production",
|
|
1717
2216
|
capabilities: {
|
|
1718
|
-
supportsImages: true
|
|
2217
|
+
supportsImages: true,
|
|
2218
|
+
supportsAudio: false,
|
|
2219
|
+
supportsTranscription: false,
|
|
2220
|
+
supportsSearch: false
|
|
1719
2221
|
}
|
|
1720
2222
|
},
|
|
1721
2223
|
"xai:grok-3-mini": {
|
|
@@ -1733,7 +2235,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1733
2235
|
tags: ["low-cost", "general-purpose"],
|
|
1734
2236
|
lifecycle: "production",
|
|
1735
2237
|
capabilities: {
|
|
1736
|
-
supportsImages: false
|
|
2238
|
+
supportsImages: false,
|
|
2239
|
+
supportsAudio: false,
|
|
2240
|
+
supportsTranscription: false,
|
|
2241
|
+
supportsSearch: false
|
|
1737
2242
|
}
|
|
1738
2243
|
},
|
|
1739
2244
|
"xai:grok-3": {
|
|
@@ -1751,7 +2256,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1751
2256
|
tags: ["general-purpose"],
|
|
1752
2257
|
lifecycle: "production",
|
|
1753
2258
|
capabilities: {
|
|
1754
|
-
supportsImages: true
|
|
2259
|
+
supportsImages: true,
|
|
2260
|
+
supportsAudio: false,
|
|
2261
|
+
supportsTranscription: false,
|
|
2262
|
+
supportsSearch: false
|
|
1755
2263
|
}
|
|
1756
2264
|
},
|
|
1757
2265
|
"openrouter:gpt-oss-120b": {
|
|
@@ -1769,9 +2277,34 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1769
2277
|
tags: ["preview", "general-purpose", "reasoning"],
|
|
1770
2278
|
lifecycle: "production",
|
|
1771
2279
|
capabilities: {
|
|
1772
|
-
supportsImages: false
|
|
2280
|
+
supportsImages: false,
|
|
2281
|
+
supportsAudio: false,
|
|
2282
|
+
supportsTranscription: false,
|
|
2283
|
+
supportsSearch: false
|
|
1773
2284
|
}
|
|
1774
2285
|
},
|
|
2286
|
+
"fireworks-ai:qwen3-8b": {
|
|
2287
|
+
id: "fireworks-ai:qwen3-8b",
|
|
2288
|
+
name: "Qwen3 8B",
|
|
2289
|
+
description: "Qwen3 8B is a newer-generation small model with better architecture than Llama 3.1 8B. Same price tier with improved quality across reasoning and coding tasks.",
|
|
2290
|
+
input: {
|
|
2291
|
+
maxTokens: 40960,
|
|
2292
|
+
costPer1MTokens: 0.2
|
|
2293
|
+
},
|
|
2294
|
+
output: {
|
|
2295
|
+
maxTokens: 16384,
|
|
2296
|
+
costPer1MTokens: 0.2
|
|
2297
|
+
},
|
|
2298
|
+
tags: ["low-cost", "general-purpose", "reasoning"],
|
|
2299
|
+
lifecycle: "production",
|
|
2300
|
+
capabilities: {
|
|
2301
|
+
supportsImages: false,
|
|
2302
|
+
supportsAudio: false,
|
|
2303
|
+
supportsTranscription: false,
|
|
2304
|
+
supportsSearch: false
|
|
2305
|
+
},
|
|
2306
|
+
aliases: ["accounts/fireworks/models/qwen3-8b"]
|
|
2307
|
+
},
|
|
1775
2308
|
"fireworks-ai:gpt-oss-20b": {
|
|
1776
2309
|
id: "fireworks-ai:gpt-oss-20b",
|
|
1777
2310
|
name: "GPT-OSS 20B",
|
|
@@ -1787,7 +2320,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1787
2320
|
tags: ["general-purpose", "reasoning", "low-cost"],
|
|
1788
2321
|
lifecycle: "production",
|
|
1789
2322
|
capabilities: {
|
|
1790
|
-
supportsImages: false
|
|
2323
|
+
supportsImages: false,
|
|
2324
|
+
supportsAudio: false,
|
|
2325
|
+
supportsTranscription: false,
|
|
2326
|
+
supportsSearch: false
|
|
1791
2327
|
},
|
|
1792
2328
|
aliases: ["accounts/fireworks/models/gpt-oss-20b"]
|
|
1793
2329
|
},
|
|
@@ -1806,10 +2342,57 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1806
2342
|
tags: ["general-purpose", "reasoning"],
|
|
1807
2343
|
lifecycle: "production",
|
|
1808
2344
|
capabilities: {
|
|
1809
|
-
supportsImages: false
|
|
2345
|
+
supportsImages: false,
|
|
2346
|
+
supportsAudio: false,
|
|
2347
|
+
supportsTranscription: false,
|
|
2348
|
+
supportsSearch: false
|
|
1810
2349
|
},
|
|
1811
2350
|
aliases: ["accounts/fireworks/models/gpt-oss-120b"]
|
|
1812
2351
|
},
|
|
2352
|
+
"fireworks-ai:deepseek-v3p2": {
|
|
2353
|
+
id: "fireworks-ai:deepseek-v3p2",
|
|
2354
|
+
name: "DeepSeek V3.2",
|
|
2355
|
+
description: "DeepSeek V3.2 is a 675B-parameter mixture-of-experts model with superior reasoning and agent performance. It delivers high computational efficiency with strong results across coding, math, and general-purpose tasks.",
|
|
2356
|
+
input: {
|
|
2357
|
+
maxTokens: 163840,
|
|
2358
|
+
costPer1MTokens: 0.56
|
|
2359
|
+
},
|
|
2360
|
+
output: {
|
|
2361
|
+
maxTokens: 16e4,
|
|
2362
|
+
costPer1MTokens: 1.68
|
|
2363
|
+
},
|
|
2364
|
+
tags: ["recommended", "reasoning", "general-purpose", "coding"],
|
|
2365
|
+
lifecycle: "production",
|
|
2366
|
+
capabilities: {
|
|
2367
|
+
supportsImages: false,
|
|
2368
|
+
supportsAudio: false,
|
|
2369
|
+
supportsTranscription: false,
|
|
2370
|
+
supportsSearch: false
|
|
2371
|
+
},
|
|
2372
|
+
aliases: ["accounts/fireworks/models/deepseek-v3p2"]
|
|
2373
|
+
},
|
|
2374
|
+
"fireworks-ai:deepseek-v3p1": {
|
|
2375
|
+
id: "fireworks-ai:deepseek-v3p1",
|
|
2376
|
+
name: "DeepSeek V3.1",
|
|
2377
|
+
description: "DeepSeek V3.1 is a 685B-parameter hybrid LLM with mixture-of-experts architecture (37B activated per token). Features thinking and non-thinking chat modes for complex agentic behaviors and reasoning tasks.",
|
|
2378
|
+
input: {
|
|
2379
|
+
maxTokens: 163840,
|
|
2380
|
+
costPer1MTokens: 0.56
|
|
2381
|
+
},
|
|
2382
|
+
output: {
|
|
2383
|
+
maxTokens: 163840,
|
|
2384
|
+
costPer1MTokens: 1.68
|
|
2385
|
+
},
|
|
2386
|
+
tags: ["reasoning", "general-purpose", "coding"],
|
|
2387
|
+
lifecycle: "production",
|
|
2388
|
+
capabilities: {
|
|
2389
|
+
supportsImages: false,
|
|
2390
|
+
supportsAudio: false,
|
|
2391
|
+
supportsTranscription: false,
|
|
2392
|
+
supportsSearch: false
|
|
2393
|
+
},
|
|
2394
|
+
aliases: ["accounts/fireworks/models/deepseek-v3p1"]
|
|
2395
|
+
},
|
|
1813
2396
|
"fireworks-ai:deepseek-r1-0528": {
|
|
1814
2397
|
id: "fireworks-ai:deepseek-r1-0528",
|
|
1815
2398
|
name: "DeepSeek R1 0528",
|
|
@@ -1823,9 +2406,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1823
2406
|
costPer1MTokens: 8
|
|
1824
2407
|
},
|
|
1825
2408
|
tags: ["recommended", "reasoning", "general-purpose", "coding"],
|
|
1826
|
-
lifecycle: "
|
|
2409
|
+
lifecycle: "discontinued",
|
|
1827
2410
|
capabilities: {
|
|
1828
|
-
supportsImages: false
|
|
2411
|
+
supportsImages: false,
|
|
2412
|
+
supportsAudio: false,
|
|
2413
|
+
supportsTranscription: false,
|
|
2414
|
+
supportsSearch: false
|
|
1829
2415
|
},
|
|
1830
2416
|
aliases: ["accounts/fireworks/models/deepseek-r1-0528"]
|
|
1831
2417
|
},
|
|
@@ -1842,9 +2428,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1842
2428
|
costPer1MTokens: 0.9
|
|
1843
2429
|
},
|
|
1844
2430
|
tags: ["recommended", "general-purpose"],
|
|
1845
|
-
lifecycle: "
|
|
2431
|
+
lifecycle: "discontinued",
|
|
1846
2432
|
capabilities: {
|
|
1847
|
-
supportsImages: false
|
|
2433
|
+
supportsImages: false,
|
|
2434
|
+
supportsAudio: false,
|
|
2435
|
+
supportsTranscription: false,
|
|
2436
|
+
supportsSearch: false
|
|
1848
2437
|
},
|
|
1849
2438
|
aliases: ["accounts/fireworks/models/deepseek-v3-0324"]
|
|
1850
2439
|
},
|
|
@@ -1861,9 +2450,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1861
2450
|
costPer1MTokens: 0.88
|
|
1862
2451
|
},
|
|
1863
2452
|
tags: ["general-purpose", "vision"],
|
|
1864
|
-
lifecycle: "
|
|
2453
|
+
lifecycle: "discontinued",
|
|
1865
2454
|
capabilities: {
|
|
1866
|
-
supportsImages: true
|
|
2455
|
+
supportsImages: true,
|
|
2456
|
+
supportsAudio: false,
|
|
2457
|
+
supportsTranscription: false,
|
|
2458
|
+
supportsSearch: false
|
|
1867
2459
|
},
|
|
1868
2460
|
aliases: ["accounts/fireworks/models/llama4-maverick-instruct-basic"]
|
|
1869
2461
|
},
|
|
@@ -1880,9 +2472,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1880
2472
|
costPer1MTokens: 0.6
|
|
1881
2473
|
},
|
|
1882
2474
|
tags: ["general-purpose", "vision"],
|
|
1883
|
-
lifecycle: "
|
|
2475
|
+
lifecycle: "discontinued",
|
|
1884
2476
|
capabilities: {
|
|
1885
|
-
supportsImages: true
|
|
2477
|
+
supportsImages: true,
|
|
2478
|
+
supportsAudio: false,
|
|
2479
|
+
supportsTranscription: false,
|
|
2480
|
+
supportsSearch: false
|
|
1886
2481
|
},
|
|
1887
2482
|
aliases: ["accounts/fireworks/models/llama4-scout-instruct-basic"]
|
|
1888
2483
|
},
|
|
@@ -1901,7 +2496,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1901
2496
|
tags: ["general-purpose"],
|
|
1902
2497
|
lifecycle: "production",
|
|
1903
2498
|
capabilities: {
|
|
1904
|
-
supportsImages: false
|
|
2499
|
+
supportsImages: false,
|
|
2500
|
+
supportsAudio: false,
|
|
2501
|
+
supportsTranscription: false,
|
|
2502
|
+
supportsSearch: false
|
|
1905
2503
|
},
|
|
1906
2504
|
aliases: ["accounts/fireworks/models/llama-v3p3-70b-instruct"]
|
|
1907
2505
|
},
|
|
@@ -1920,7 +2518,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1920
2518
|
tags: ["reasoning", "general-purpose", "coding"],
|
|
1921
2519
|
lifecycle: "discontinued",
|
|
1922
2520
|
capabilities: {
|
|
1923
|
-
supportsImages: false
|
|
2521
|
+
supportsImages: false,
|
|
2522
|
+
supportsAudio: false,
|
|
2523
|
+
supportsTranscription: false,
|
|
2524
|
+
supportsSearch: false
|
|
1924
2525
|
},
|
|
1925
2526
|
aliases: ["accounts/fireworks/models/deepseek-r1"]
|
|
1926
2527
|
},
|
|
@@ -1939,7 +2540,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1939
2540
|
tags: ["reasoning", "general-purpose", "coding"],
|
|
1940
2541
|
lifecycle: "discontinued",
|
|
1941
2542
|
capabilities: {
|
|
1942
|
-
supportsImages: false
|
|
2543
|
+
supportsImages: false,
|
|
2544
|
+
supportsAudio: false,
|
|
2545
|
+
supportsTranscription: false,
|
|
2546
|
+
supportsSearch: false
|
|
1943
2547
|
},
|
|
1944
2548
|
aliases: ["accounts/fireworks/models/deepseek-r1-basic"]
|
|
1945
2549
|
},
|
|
@@ -1958,7 +2562,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1958
2562
|
tags: ["deprecated", "general-purpose"],
|
|
1959
2563
|
lifecycle: "discontinued",
|
|
1960
2564
|
capabilities: {
|
|
1961
|
-
supportsImages: false
|
|
2565
|
+
supportsImages: false,
|
|
2566
|
+
supportsAudio: false,
|
|
2567
|
+
supportsTranscription: false,
|
|
2568
|
+
supportsSearch: false
|
|
1962
2569
|
},
|
|
1963
2570
|
aliases: ["accounts/fireworks/models/deepseek-v3"]
|
|
1964
2571
|
},
|
|
@@ -1977,7 +2584,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1977
2584
|
tags: ["deprecated", "general-purpose"],
|
|
1978
2585
|
lifecycle: "discontinued",
|
|
1979
2586
|
capabilities: {
|
|
1980
|
-
supportsImages: false
|
|
2587
|
+
supportsImages: false,
|
|
2588
|
+
supportsAudio: false,
|
|
2589
|
+
supportsTranscription: false,
|
|
2590
|
+
supportsSearch: false
|
|
1981
2591
|
},
|
|
1982
2592
|
aliases: ["accounts/fireworks/models/llama-v3p1-405b-instruct"]
|
|
1983
2593
|
},
|
|
@@ -1996,7 +2606,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1996
2606
|
tags: ["deprecated", "general-purpose"],
|
|
1997
2607
|
lifecycle: "discontinued",
|
|
1998
2608
|
capabilities: {
|
|
1999
|
-
supportsImages: false
|
|
2609
|
+
supportsImages: false,
|
|
2610
|
+
supportsAudio: false,
|
|
2611
|
+
supportsTranscription: false,
|
|
2612
|
+
supportsSearch: false
|
|
2000
2613
|
},
|
|
2001
2614
|
aliases: ["accounts/fireworks/models/llama-v3p1-70b-instruct"]
|
|
2002
2615
|
},
|
|
@@ -2013,9 +2626,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2013
2626
|
costPer1MTokens: 0.2
|
|
2014
2627
|
},
|
|
2015
2628
|
tags: ["low-cost", "general-purpose"],
|
|
2016
|
-
lifecycle: "
|
|
2629
|
+
lifecycle: "discontinued",
|
|
2017
2630
|
capabilities: {
|
|
2018
|
-
supportsImages: false
|
|
2631
|
+
supportsImages: false,
|
|
2632
|
+
supportsAudio: false,
|
|
2633
|
+
supportsTranscription: false,
|
|
2634
|
+
supportsSearch: false
|
|
2019
2635
|
},
|
|
2020
2636
|
aliases: ["accounts/fireworks/models/llama-v3p1-8b-instruct"]
|
|
2021
2637
|
},
|
|
@@ -2034,7 +2650,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2034
2650
|
tags: ["deprecated", "general-purpose"],
|
|
2035
2651
|
lifecycle: "discontinued",
|
|
2036
2652
|
capabilities: {
|
|
2037
|
-
supportsImages: false
|
|
2653
|
+
supportsImages: false,
|
|
2654
|
+
supportsAudio: false,
|
|
2655
|
+
supportsTranscription: false,
|
|
2656
|
+
supportsSearch: false
|
|
2038
2657
|
},
|
|
2039
2658
|
aliases: ["accounts/fireworks/models/mixtral-8x22b-instruct"]
|
|
2040
2659
|
},
|
|
@@ -2053,7 +2672,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2053
2672
|
tags: ["low-cost", "general-purpose"],
|
|
2054
2673
|
lifecycle: "discontinued",
|
|
2055
2674
|
capabilities: {
|
|
2056
|
-
supportsImages: false
|
|
2675
|
+
supportsImages: false,
|
|
2676
|
+
supportsAudio: false,
|
|
2677
|
+
supportsTranscription: false,
|
|
2678
|
+
supportsSearch: false
|
|
2057
2679
|
},
|
|
2058
2680
|
aliases: ["accounts/fireworks/models/mixtral-8x7b-instruct"]
|
|
2059
2681
|
},
|
|
@@ -2072,7 +2694,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2072
2694
|
tags: ["roleplay", "storytelling", "low-cost"],
|
|
2073
2695
|
lifecycle: "discontinued",
|
|
2074
2696
|
capabilities: {
|
|
2075
|
-
supportsImages: false
|
|
2697
|
+
supportsImages: false,
|
|
2698
|
+
supportsAudio: false,
|
|
2699
|
+
supportsTranscription: false,
|
|
2700
|
+
supportsSearch: false
|
|
2076
2701
|
},
|
|
2077
2702
|
aliases: ["accounts/fireworks/models/mythomax-l2-13b"]
|
|
2078
2703
|
},
|
|
@@ -2091,9 +2716,34 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2091
2716
|
tags: ["deprecated", "low-cost", "general-purpose"],
|
|
2092
2717
|
lifecycle: "discontinued",
|
|
2093
2718
|
capabilities: {
|
|
2094
|
-
supportsImages: false
|
|
2719
|
+
supportsImages: false,
|
|
2720
|
+
supportsAudio: false,
|
|
2721
|
+
supportsTranscription: false,
|
|
2722
|
+
supportsSearch: false
|
|
2095
2723
|
},
|
|
2096
2724
|
aliases: ["accounts/fireworks/models/gemma2-9b-it"]
|
|
2725
|
+
},
|
|
2726
|
+
"fireworks-ai:whisper-v3": {
|
|
2727
|
+
id: "fireworks-ai:whisper-v3",
|
|
2728
|
+
name: "Whisper V3",
|
|
2729
|
+
description: "Whisper V3 on Fireworks AI \u2014 multilingual speech recognition with high accuracy.",
|
|
2730
|
+
input: {
|
|
2731
|
+
maxTokens: 0,
|
|
2732
|
+
costPer1MTokens: 0,
|
|
2733
|
+
costPerMinute: 15e-4
|
|
2734
|
+
},
|
|
2735
|
+
output: {
|
|
2736
|
+
maxTokens: 0,
|
|
2737
|
+
costPer1MTokens: 0
|
|
2738
|
+
},
|
|
2739
|
+
tags: ["general-purpose", "speech-to-text"],
|
|
2740
|
+
lifecycle: "production",
|
|
2741
|
+
capabilities: {
|
|
2742
|
+
supportsImages: false,
|
|
2743
|
+
supportsAudio: false,
|
|
2744
|
+
supportsTranscription: true,
|
|
2745
|
+
supportsSearch: false
|
|
2746
|
+
}
|
|
2097
2747
|
}
|
|
2098
2748
|
};
|
|
2099
2749
|
var defaultModel = {
|
|
@@ -2157,7 +2807,7 @@ var CognitiveBeta = class _CognitiveBeta {
|
|
|
2157
2807
|
}
|
|
2158
2808
|
async generateText(input, options = {}) {
|
|
2159
2809
|
const signal = options.signal ?? AbortSignal.timeout(this._timeout);
|
|
2160
|
-
const req = { input };
|
|
2810
|
+
const req = { type: "generateText", input };
|
|
2161
2811
|
this._events.emit("request", req);
|
|
2162
2812
|
try {
|
|
2163
2813
|
const { data } = await this._withServerRetry(
|
|
@@ -2181,9 +2831,32 @@ var CognitiveBeta = class _CognitiveBeta {
|
|
|
2181
2831
|
);
|
|
2182
2832
|
return data.models;
|
|
2183
2833
|
}
|
|
2834
|
+
async transcribeAudio(input, options = {}) {
|
|
2835
|
+
const signal = options.signal ?? AbortSignal.timeout(this._timeout);
|
|
2836
|
+
const req = { type: "transcribeAudio", input };
|
|
2837
|
+
this._events.emit("request", req);
|
|
2838
|
+
try {
|
|
2839
|
+
const { data } = await this._withServerRetry(
|
|
2840
|
+
() => this._axiosClient.post("/v2/cognitive/transcribe-audio", input, {
|
|
2841
|
+
signal,
|
|
2842
|
+
timeout: options.timeout ?? this._timeout
|
|
2843
|
+
}),
|
|
2844
|
+
options,
|
|
2845
|
+
req
|
|
2846
|
+
);
|
|
2847
|
+
if (data.error) {
|
|
2848
|
+
throw new Error(`Transcription error: ${data.error}`);
|
|
2849
|
+
}
|
|
2850
|
+
this._events.emit("response", req, data);
|
|
2851
|
+
return data;
|
|
2852
|
+
} catch (error) {
|
|
2853
|
+
this._events.emit("error", req, error);
|
|
2854
|
+
throw error;
|
|
2855
|
+
}
|
|
2856
|
+
}
|
|
2184
2857
|
async *generateTextStream(request, options = {}) {
|
|
2185
2858
|
const signal = options.signal ?? AbortSignal.timeout(this._timeout);
|
|
2186
|
-
const req = { input: request };
|
|
2859
|
+
const req = { type: "generateText", input: request };
|
|
2187
2860
|
const chunks = [];
|
|
2188
2861
|
let lastChunk;
|
|
2189
2862
|
this._events.emit("request", req);
|