@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.cjs
CHANGED
|
@@ -663,10 +663,79 @@ var models = {
|
|
|
663
663
|
tags: ["recommended", "reasoning", "general-purpose", "vision", "coding", "agents"],
|
|
664
664
|
lifecycle: "production",
|
|
665
665
|
capabilities: {
|
|
666
|
-
supportsImages: true
|
|
666
|
+
supportsImages: true,
|
|
667
|
+
supportsAudio: false,
|
|
668
|
+
supportsTranscription: false,
|
|
669
|
+
supportsSearch: true
|
|
667
670
|
},
|
|
668
671
|
aliases: ["gpt-5.4"]
|
|
669
672
|
},
|
|
673
|
+
"openai:gpt-5.4-mini-2026-03-17": {
|
|
674
|
+
id: "openai:gpt-5.4-mini-2026-03-17",
|
|
675
|
+
name: "GPT-5.4 Mini",
|
|
676
|
+
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.",
|
|
677
|
+
input: {
|
|
678
|
+
maxTokens: 4e5,
|
|
679
|
+
costPer1MTokens: 0.75
|
|
680
|
+
},
|
|
681
|
+
output: {
|
|
682
|
+
maxTokens: 128e3,
|
|
683
|
+
costPer1MTokens: 4.5
|
|
684
|
+
},
|
|
685
|
+
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
686
|
+
lifecycle: "production",
|
|
687
|
+
capabilities: {
|
|
688
|
+
supportsImages: true,
|
|
689
|
+
supportsAudio: false,
|
|
690
|
+
supportsTranscription: false,
|
|
691
|
+
supportsSearch: false
|
|
692
|
+
},
|
|
693
|
+
aliases: ["gpt-5.4-mini"]
|
|
694
|
+
},
|
|
695
|
+
"openai:gpt-5.4-nano-2026-03-17": {
|
|
696
|
+
id: "openai:gpt-5.4-nano-2026-03-17",
|
|
697
|
+
name: "GPT-5.4 Nano",
|
|
698
|
+
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.",
|
|
699
|
+
input: {
|
|
700
|
+
maxTokens: 4e5,
|
|
701
|
+
costPer1MTokens: 0.2
|
|
702
|
+
},
|
|
703
|
+
output: {
|
|
704
|
+
maxTokens: 128e3,
|
|
705
|
+
costPer1MTokens: 1.25
|
|
706
|
+
},
|
|
707
|
+
tags: ["low-cost", "reasoning", "general-purpose", "vision"],
|
|
708
|
+
lifecycle: "production",
|
|
709
|
+
capabilities: {
|
|
710
|
+
supportsImages: true,
|
|
711
|
+
supportsAudio: false,
|
|
712
|
+
supportsTranscription: false,
|
|
713
|
+
supportsSearch: false
|
|
714
|
+
},
|
|
715
|
+
aliases: ["gpt-5.4-nano"]
|
|
716
|
+
},
|
|
717
|
+
"openai:gpt-5.3-chat": {
|
|
718
|
+
id: "openai:gpt-5.3-chat",
|
|
719
|
+
name: "GPT-5.3 Chat",
|
|
720
|
+
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.",
|
|
721
|
+
input: {
|
|
722
|
+
maxTokens: 128e3,
|
|
723
|
+
costPer1MTokens: 1.75
|
|
724
|
+
},
|
|
725
|
+
output: {
|
|
726
|
+
maxTokens: 16384,
|
|
727
|
+
costPer1MTokens: 14
|
|
728
|
+
},
|
|
729
|
+
tags: ["reasoning", "general-purpose", "vision"],
|
|
730
|
+
lifecycle: "preview",
|
|
731
|
+
capabilities: {
|
|
732
|
+
supportsImages: true,
|
|
733
|
+
supportsAudio: false,
|
|
734
|
+
supportsTranscription: false,
|
|
735
|
+
supportsSearch: false
|
|
736
|
+
},
|
|
737
|
+
aliases: ["gpt-5.3-chat-latest"]
|
|
738
|
+
},
|
|
670
739
|
"openai:gpt-5.2-2025-12-11": {
|
|
671
740
|
id: "openai:gpt-5.2-2025-12-11",
|
|
672
741
|
name: "GPT-5.2",
|
|
@@ -682,7 +751,10 @@ var models = {
|
|
|
682
751
|
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
683
752
|
lifecycle: "production",
|
|
684
753
|
capabilities: {
|
|
685
|
-
supportsImages: true
|
|
754
|
+
supportsImages: true,
|
|
755
|
+
supportsAudio: false,
|
|
756
|
+
supportsTranscription: false,
|
|
757
|
+
supportsSearch: true
|
|
686
758
|
}
|
|
687
759
|
},
|
|
688
760
|
"openai:gpt-5.1-2025-11-13": {
|
|
@@ -700,7 +772,10 @@ var models = {
|
|
|
700
772
|
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
701
773
|
lifecycle: "production",
|
|
702
774
|
capabilities: {
|
|
703
|
-
supportsImages: true
|
|
775
|
+
supportsImages: true,
|
|
776
|
+
supportsAudio: false,
|
|
777
|
+
supportsTranscription: false,
|
|
778
|
+
supportsSearch: true
|
|
704
779
|
}
|
|
705
780
|
},
|
|
706
781
|
"openai:gpt-5-2025-08-07": {
|
|
@@ -718,7 +793,10 @@ var models = {
|
|
|
718
793
|
tags: ["reasoning", "general-purpose", "vision"],
|
|
719
794
|
lifecycle: "production",
|
|
720
795
|
capabilities: {
|
|
721
|
-
supportsImages: true
|
|
796
|
+
supportsImages: true,
|
|
797
|
+
supportsAudio: false,
|
|
798
|
+
supportsTranscription: false,
|
|
799
|
+
supportsSearch: false
|
|
722
800
|
},
|
|
723
801
|
aliases: ["gpt-5"]
|
|
724
802
|
},
|
|
@@ -737,7 +815,10 @@ var models = {
|
|
|
737
815
|
tags: ["recommended", "reasoning", "general-purpose", "vision"],
|
|
738
816
|
lifecycle: "production",
|
|
739
817
|
capabilities: {
|
|
740
|
-
supportsImages: true
|
|
818
|
+
supportsImages: true,
|
|
819
|
+
supportsAudio: false,
|
|
820
|
+
supportsTranscription: false,
|
|
821
|
+
supportsSearch: false
|
|
741
822
|
},
|
|
742
823
|
aliases: ["gpt-5-mini"]
|
|
743
824
|
},
|
|
@@ -756,7 +837,10 @@ var models = {
|
|
|
756
837
|
tags: ["low-cost", "reasoning", "general-purpose", "vision"],
|
|
757
838
|
lifecycle: "production",
|
|
758
839
|
capabilities: {
|
|
759
|
-
supportsImages: true
|
|
840
|
+
supportsImages: true,
|
|
841
|
+
supportsAudio: false,
|
|
842
|
+
supportsTranscription: false,
|
|
843
|
+
supportsSearch: false
|
|
760
844
|
},
|
|
761
845
|
aliases: ["gpt-5-nano"]
|
|
762
846
|
},
|
|
@@ -775,7 +859,10 @@ var models = {
|
|
|
775
859
|
tags: ["reasoning", "vision", "coding"],
|
|
776
860
|
lifecycle: "production",
|
|
777
861
|
capabilities: {
|
|
778
|
-
supportsImages: true
|
|
862
|
+
supportsImages: true,
|
|
863
|
+
supportsAudio: false,
|
|
864
|
+
supportsTranscription: false,
|
|
865
|
+
supportsSearch: false
|
|
779
866
|
},
|
|
780
867
|
aliases: ["o4-mini"]
|
|
781
868
|
},
|
|
@@ -794,7 +881,10 @@ var models = {
|
|
|
794
881
|
tags: ["reasoning", "vision", "coding"],
|
|
795
882
|
lifecycle: "production",
|
|
796
883
|
capabilities: {
|
|
797
|
-
supportsImages: true
|
|
884
|
+
supportsImages: true,
|
|
885
|
+
supportsAudio: false,
|
|
886
|
+
supportsTranscription: false,
|
|
887
|
+
supportsSearch: false
|
|
798
888
|
},
|
|
799
889
|
aliases: ["o3"]
|
|
800
890
|
},
|
|
@@ -813,7 +903,10 @@ var models = {
|
|
|
813
903
|
tags: ["recommended", "vision", "general-purpose"],
|
|
814
904
|
lifecycle: "production",
|
|
815
905
|
capabilities: {
|
|
816
|
-
supportsImages: true
|
|
906
|
+
supportsImages: true,
|
|
907
|
+
supportsAudio: false,
|
|
908
|
+
supportsTranscription: false,
|
|
909
|
+
supportsSearch: false
|
|
817
910
|
},
|
|
818
911
|
aliases: ["gpt-4.1"]
|
|
819
912
|
},
|
|
@@ -832,7 +925,10 @@ var models = {
|
|
|
832
925
|
tags: ["recommended", "vision", "general-purpose"],
|
|
833
926
|
lifecycle: "production",
|
|
834
927
|
capabilities: {
|
|
835
|
-
supportsImages: true
|
|
928
|
+
supportsImages: true,
|
|
929
|
+
supportsAudio: false,
|
|
930
|
+
supportsTranscription: false,
|
|
931
|
+
supportsSearch: false
|
|
836
932
|
},
|
|
837
933
|
aliases: ["gpt-4.1-mini"]
|
|
838
934
|
},
|
|
@@ -851,7 +947,10 @@ var models = {
|
|
|
851
947
|
tags: ["low-cost", "vision", "general-purpose"],
|
|
852
948
|
lifecycle: "production",
|
|
853
949
|
capabilities: {
|
|
854
|
-
supportsImages: true
|
|
950
|
+
supportsImages: true,
|
|
951
|
+
supportsAudio: false,
|
|
952
|
+
supportsTranscription: false,
|
|
953
|
+
supportsSearch: false
|
|
855
954
|
},
|
|
856
955
|
aliases: ["gpt-4.1-nano"]
|
|
857
956
|
},
|
|
@@ -870,7 +969,10 @@ var models = {
|
|
|
870
969
|
tags: ["reasoning", "general-purpose", "coding"],
|
|
871
970
|
lifecycle: "production",
|
|
872
971
|
capabilities: {
|
|
873
|
-
supportsImages: false
|
|
972
|
+
supportsImages: false,
|
|
973
|
+
supportsAudio: false,
|
|
974
|
+
supportsTranscription: false,
|
|
975
|
+
supportsSearch: false
|
|
874
976
|
},
|
|
875
977
|
aliases: ["o3-mini"]
|
|
876
978
|
},
|
|
@@ -889,7 +991,10 @@ var models = {
|
|
|
889
991
|
tags: ["reasoning", "vision", "general-purpose"],
|
|
890
992
|
lifecycle: "production",
|
|
891
993
|
capabilities: {
|
|
892
|
-
supportsImages: true
|
|
994
|
+
supportsImages: true,
|
|
995
|
+
supportsAudio: false,
|
|
996
|
+
supportsTranscription: false,
|
|
997
|
+
supportsSearch: false
|
|
893
998
|
}
|
|
894
999
|
},
|
|
895
1000
|
"openai:o1-mini-2024-09-12": {
|
|
@@ -905,9 +1010,12 @@ var models = {
|
|
|
905
1010
|
costPer1MTokens: 4.4
|
|
906
1011
|
},
|
|
907
1012
|
tags: ["reasoning", "vision", "general-purpose"],
|
|
908
|
-
lifecycle: "
|
|
1013
|
+
lifecycle: "discontinued",
|
|
909
1014
|
capabilities: {
|
|
910
|
-
supportsImages: false
|
|
1015
|
+
supportsImages: false,
|
|
1016
|
+
supportsAudio: false,
|
|
1017
|
+
supportsTranscription: false,
|
|
1018
|
+
supportsSearch: false
|
|
911
1019
|
},
|
|
912
1020
|
aliases: ["o1-mini"]
|
|
913
1021
|
},
|
|
@@ -926,7 +1034,10 @@ var models = {
|
|
|
926
1034
|
tags: ["recommended", "vision", "low-cost", "general-purpose"],
|
|
927
1035
|
lifecycle: "production",
|
|
928
1036
|
capabilities: {
|
|
929
|
-
supportsImages: true
|
|
1037
|
+
supportsImages: true,
|
|
1038
|
+
supportsAudio: false,
|
|
1039
|
+
supportsTranscription: false,
|
|
1040
|
+
supportsSearch: false
|
|
930
1041
|
},
|
|
931
1042
|
aliases: ["gpt-4o-mini"]
|
|
932
1043
|
},
|
|
@@ -945,7 +1056,10 @@ var models = {
|
|
|
945
1056
|
tags: ["recommended", "vision", "general-purpose", "coding", "agents"],
|
|
946
1057
|
lifecycle: "production",
|
|
947
1058
|
capabilities: {
|
|
948
|
-
supportsImages: true
|
|
1059
|
+
supportsImages: true,
|
|
1060
|
+
supportsAudio: false,
|
|
1061
|
+
supportsTranscription: false,
|
|
1062
|
+
supportsSearch: false
|
|
949
1063
|
},
|
|
950
1064
|
aliases: ["gpt-4o"]
|
|
951
1065
|
},
|
|
@@ -964,7 +1078,10 @@ var models = {
|
|
|
964
1078
|
tags: ["deprecated", "vision", "general-purpose", "coding", "agents"],
|
|
965
1079
|
lifecycle: "deprecated",
|
|
966
1080
|
capabilities: {
|
|
967
|
-
supportsImages: true
|
|
1081
|
+
supportsImages: true,
|
|
1082
|
+
supportsAudio: false,
|
|
1083
|
+
supportsTranscription: false,
|
|
1084
|
+
supportsSearch: false
|
|
968
1085
|
}
|
|
969
1086
|
},
|
|
970
1087
|
"openai:gpt-4o-2024-05-13": {
|
|
@@ -982,7 +1099,10 @@ var models = {
|
|
|
982
1099
|
tags: ["deprecated", "vision", "general-purpose", "coding", "agents"],
|
|
983
1100
|
lifecycle: "deprecated",
|
|
984
1101
|
capabilities: {
|
|
985
|
-
supportsImages: true
|
|
1102
|
+
supportsImages: true,
|
|
1103
|
+
supportsAudio: false,
|
|
1104
|
+
supportsTranscription: false,
|
|
1105
|
+
supportsSearch: false
|
|
986
1106
|
}
|
|
987
1107
|
},
|
|
988
1108
|
"openai:gpt-4-turbo-2024-04-09": {
|
|
@@ -1000,7 +1120,10 @@ var models = {
|
|
|
1000
1120
|
tags: ["deprecated", "general-purpose", "coding", "agents"],
|
|
1001
1121
|
lifecycle: "deprecated",
|
|
1002
1122
|
capabilities: {
|
|
1003
|
-
supportsImages: true
|
|
1123
|
+
supportsImages: true,
|
|
1124
|
+
supportsAudio: false,
|
|
1125
|
+
supportsTranscription: false,
|
|
1126
|
+
supportsSearch: false
|
|
1004
1127
|
}
|
|
1005
1128
|
},
|
|
1006
1129
|
"openai:gpt-3.5-turbo-0125": {
|
|
@@ -1018,7 +1141,32 @@ var models = {
|
|
|
1018
1141
|
tags: ["deprecated", "general-purpose", "low-cost"],
|
|
1019
1142
|
lifecycle: "deprecated",
|
|
1020
1143
|
capabilities: {
|
|
1021
|
-
supportsImages: false
|
|
1144
|
+
supportsImages: false,
|
|
1145
|
+
supportsAudio: false,
|
|
1146
|
+
supportsTranscription: false,
|
|
1147
|
+
supportsSearch: false
|
|
1148
|
+
}
|
|
1149
|
+
},
|
|
1150
|
+
"openai:whisper-1": {
|
|
1151
|
+
id: "openai:whisper-1",
|
|
1152
|
+
name: "Whisper V2",
|
|
1153
|
+
description: "OpenAI Whisper V2 \u2014 general-purpose speech recognition model supporting 99 languages.",
|
|
1154
|
+
input: {
|
|
1155
|
+
maxTokens: 0,
|
|
1156
|
+
costPer1MTokens: 0,
|
|
1157
|
+
costPerMinute: 6e-3
|
|
1158
|
+
},
|
|
1159
|
+
output: {
|
|
1160
|
+
maxTokens: 0,
|
|
1161
|
+
costPer1MTokens: 0
|
|
1162
|
+
},
|
|
1163
|
+
tags: ["general-purpose", "speech-to-text"],
|
|
1164
|
+
lifecycle: "production",
|
|
1165
|
+
capabilities: {
|
|
1166
|
+
supportsImages: false,
|
|
1167
|
+
supportsAudio: false,
|
|
1168
|
+
supportsTranscription: true,
|
|
1169
|
+
supportsSearch: false
|
|
1022
1170
|
}
|
|
1023
1171
|
},
|
|
1024
1172
|
"anthropic:claude-opus-4-6": {
|
|
@@ -1026,7 +1174,7 @@ var models = {
|
|
|
1026
1174
|
name: "Claude Opus 4.6",
|
|
1027
1175
|
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.",
|
|
1028
1176
|
input: {
|
|
1029
|
-
maxTokens:
|
|
1177
|
+
maxTokens: 1e6,
|
|
1030
1178
|
costPer1MTokens: 5
|
|
1031
1179
|
},
|
|
1032
1180
|
output: {
|
|
@@ -1036,7 +1184,10 @@ var models = {
|
|
|
1036
1184
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1037
1185
|
lifecycle: "production",
|
|
1038
1186
|
capabilities: {
|
|
1039
|
-
supportsImages: true
|
|
1187
|
+
supportsImages: true,
|
|
1188
|
+
supportsAudio: false,
|
|
1189
|
+
supportsTranscription: false,
|
|
1190
|
+
supportsSearch: true
|
|
1040
1191
|
}
|
|
1041
1192
|
},
|
|
1042
1193
|
"anthropic:claude-sonnet-4-6": {
|
|
@@ -1044,7 +1195,7 @@ var models = {
|
|
|
1044
1195
|
name: "Claude Sonnet 4.6",
|
|
1045
1196
|
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.",
|
|
1046
1197
|
input: {
|
|
1047
|
-
maxTokens:
|
|
1198
|
+
maxTokens: 1e6,
|
|
1048
1199
|
costPer1MTokens: 3
|
|
1049
1200
|
},
|
|
1050
1201
|
output: {
|
|
@@ -1054,9 +1205,34 @@ var models = {
|
|
|
1054
1205
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1055
1206
|
lifecycle: "production",
|
|
1056
1207
|
capabilities: {
|
|
1057
|
-
supportsImages: true
|
|
1208
|
+
supportsImages: true,
|
|
1209
|
+
supportsAudio: false,
|
|
1210
|
+
supportsTranscription: false,
|
|
1211
|
+
supportsSearch: true
|
|
1058
1212
|
}
|
|
1059
1213
|
},
|
|
1214
|
+
"anthropic:claude-opus-4-5-20251101": {
|
|
1215
|
+
id: "anthropic:claude-opus-4-5-20251101",
|
|
1216
|
+
name: "Claude Opus 4.5",
|
|
1217
|
+
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.",
|
|
1218
|
+
input: {
|
|
1219
|
+
maxTokens: 2e5,
|
|
1220
|
+
costPer1MTokens: 5
|
|
1221
|
+
},
|
|
1222
|
+
output: {
|
|
1223
|
+
maxTokens: 64e3,
|
|
1224
|
+
costPer1MTokens: 25
|
|
1225
|
+
},
|
|
1226
|
+
tags: ["reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1227
|
+
lifecycle: "production",
|
|
1228
|
+
capabilities: {
|
|
1229
|
+
supportsImages: true,
|
|
1230
|
+
supportsAudio: false,
|
|
1231
|
+
supportsTranscription: false,
|
|
1232
|
+
supportsSearch: true
|
|
1233
|
+
},
|
|
1234
|
+
aliases: ["claude-opus-4-5"]
|
|
1235
|
+
},
|
|
1060
1236
|
"anthropic:claude-sonnet-4-5-20250929": {
|
|
1061
1237
|
id: "anthropic:claude-sonnet-4-5-20250929",
|
|
1062
1238
|
name: "Claude Sonnet 4.5",
|
|
@@ -1072,7 +1248,10 @@ var models = {
|
|
|
1072
1248
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1073
1249
|
lifecycle: "production",
|
|
1074
1250
|
capabilities: {
|
|
1075
|
-
supportsImages: true
|
|
1251
|
+
supportsImages: true,
|
|
1252
|
+
supportsAudio: false,
|
|
1253
|
+
supportsTranscription: false,
|
|
1254
|
+
supportsSearch: true
|
|
1076
1255
|
},
|
|
1077
1256
|
aliases: ["claude-sonnet-4-5"]
|
|
1078
1257
|
},
|
|
@@ -1091,7 +1270,10 @@ var models = {
|
|
|
1091
1270
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1092
1271
|
lifecycle: "production",
|
|
1093
1272
|
capabilities: {
|
|
1094
|
-
supportsImages: true
|
|
1273
|
+
supportsImages: true,
|
|
1274
|
+
supportsAudio: false,
|
|
1275
|
+
supportsTranscription: false,
|
|
1276
|
+
supportsSearch: false
|
|
1095
1277
|
},
|
|
1096
1278
|
aliases: ["claude-sonnet-4"]
|
|
1097
1279
|
},
|
|
@@ -1110,7 +1292,10 @@ var models = {
|
|
|
1110
1292
|
tags: ["deprecated", "vision", "reasoning", "general-purpose", "agents", "coding"],
|
|
1111
1293
|
lifecycle: "deprecated",
|
|
1112
1294
|
capabilities: {
|
|
1113
|
-
supportsImages: true
|
|
1295
|
+
supportsImages: true,
|
|
1296
|
+
supportsAudio: false,
|
|
1297
|
+
supportsTranscription: false,
|
|
1298
|
+
supportsSearch: false
|
|
1114
1299
|
},
|
|
1115
1300
|
aliases: ["claude-sonnet-4-reasoning"]
|
|
1116
1301
|
},
|
|
@@ -1129,7 +1314,10 @@ var models = {
|
|
|
1129
1314
|
tags: ["recommended", "agents", "vision", "general-purpose", "coding"],
|
|
1130
1315
|
lifecycle: "production",
|
|
1131
1316
|
capabilities: {
|
|
1132
|
-
supportsImages: true
|
|
1317
|
+
supportsImages: true,
|
|
1318
|
+
supportsAudio: false,
|
|
1319
|
+
supportsTranscription: false,
|
|
1320
|
+
supportsSearch: false
|
|
1133
1321
|
},
|
|
1134
1322
|
aliases: ["claude-haiku-4-5"]
|
|
1135
1323
|
},
|
|
@@ -1150,9 +1338,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1150
1338
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1151
1339
|
lifecycle: "production",
|
|
1152
1340
|
capabilities: {
|
|
1153
|
-
supportsImages: true
|
|
1341
|
+
supportsImages: true,
|
|
1342
|
+
supportsAudio: false,
|
|
1343
|
+
supportsTranscription: false,
|
|
1344
|
+
supportsSearch: false
|
|
1154
1345
|
},
|
|
1155
|
-
aliases: ["claude-haiku-4-5-reasoning"]
|
|
1346
|
+
aliases: ["claude-haiku-4-5-reasoning", "claude-haiku-4-5-20251001"]
|
|
1156
1347
|
},
|
|
1157
1348
|
"anthropic:claude-3-7-sonnet-20250219": {
|
|
1158
1349
|
id: "anthropic:claude-3-7-sonnet-20250219",
|
|
@@ -1169,7 +1360,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1169
1360
|
tags: ["recommended", "reasoning", "agents", "vision", "general-purpose", "coding"],
|
|
1170
1361
|
lifecycle: "discontinued",
|
|
1171
1362
|
capabilities: {
|
|
1172
|
-
supportsImages: true
|
|
1363
|
+
supportsImages: true,
|
|
1364
|
+
supportsAudio: false,
|
|
1365
|
+
supportsTranscription: false,
|
|
1366
|
+
supportsSearch: false
|
|
1173
1367
|
}
|
|
1174
1368
|
},
|
|
1175
1369
|
"anthropic:claude-3-7-sonnet-reasoning-20250219": {
|
|
@@ -1187,7 +1381,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1187
1381
|
tags: ["deprecated", "vision", "reasoning", "general-purpose", "agents", "coding"],
|
|
1188
1382
|
lifecycle: "discontinued",
|
|
1189
1383
|
capabilities: {
|
|
1190
|
-
supportsImages: true
|
|
1384
|
+
supportsImages: true,
|
|
1385
|
+
supportsAudio: false,
|
|
1386
|
+
supportsTranscription: false,
|
|
1387
|
+
supportsSearch: false
|
|
1191
1388
|
}
|
|
1192
1389
|
},
|
|
1193
1390
|
"anthropic:claude-3-5-haiku-20241022": {
|
|
@@ -1205,7 +1402,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1205
1402
|
tags: ["general-purpose", "low-cost"],
|
|
1206
1403
|
lifecycle: "discontinued",
|
|
1207
1404
|
capabilities: {
|
|
1208
|
-
supportsImages: true
|
|
1405
|
+
supportsImages: true,
|
|
1406
|
+
supportsAudio: false,
|
|
1407
|
+
supportsTranscription: false,
|
|
1408
|
+
supportsSearch: false
|
|
1209
1409
|
}
|
|
1210
1410
|
},
|
|
1211
1411
|
"anthropic:claude-3-5-sonnet-20241022": {
|
|
@@ -1223,7 +1423,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1223
1423
|
tags: ["vision", "general-purpose", "agents", "coding", "storytelling"],
|
|
1224
1424
|
lifecycle: "discontinued",
|
|
1225
1425
|
capabilities: {
|
|
1226
|
-
supportsImages: true
|
|
1426
|
+
supportsImages: true,
|
|
1427
|
+
supportsAudio: false,
|
|
1428
|
+
supportsTranscription: false,
|
|
1429
|
+
supportsSearch: false
|
|
1227
1430
|
}
|
|
1228
1431
|
},
|
|
1229
1432
|
"anthropic:claude-3-5-sonnet-20240620": {
|
|
@@ -1241,7 +1444,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1241
1444
|
tags: ["vision", "general-purpose", "agents", "coding", "storytelling"],
|
|
1242
1445
|
lifecycle: "discontinued",
|
|
1243
1446
|
capabilities: {
|
|
1244
|
-
supportsImages: true
|
|
1447
|
+
supportsImages: true,
|
|
1448
|
+
supportsAudio: false,
|
|
1449
|
+
supportsTranscription: false,
|
|
1450
|
+
supportsSearch: false
|
|
1245
1451
|
}
|
|
1246
1452
|
},
|
|
1247
1453
|
"anthropic:claude-3-haiku-20240307": {
|
|
@@ -1259,13 +1465,16 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1259
1465
|
tags: ["deprecated", "low-cost", "general-purpose"],
|
|
1260
1466
|
lifecycle: "deprecated",
|
|
1261
1467
|
capabilities: {
|
|
1262
|
-
supportsImages: true
|
|
1468
|
+
supportsImages: true,
|
|
1469
|
+
supportsAudio: false,
|
|
1470
|
+
supportsTranscription: false,
|
|
1471
|
+
supportsSearch: false
|
|
1263
1472
|
}
|
|
1264
1473
|
},
|
|
1265
|
-
"google-ai:gemini-3-pro": {
|
|
1266
|
-
id: "google-ai:gemini-3-pro",
|
|
1267
|
-
name: "Gemini 3 Pro",
|
|
1268
|
-
description: "
|
|
1474
|
+
"google-ai:gemini-3.1-pro": {
|
|
1475
|
+
id: "google-ai:gemini-3.1-pro",
|
|
1476
|
+
name: "Gemini 3.1 Pro",
|
|
1477
|
+
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).",
|
|
1269
1478
|
input: {
|
|
1270
1479
|
maxTokens: 1048576,
|
|
1271
1480
|
costPer1MTokens: 2
|
|
@@ -1274,10 +1483,13 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1274
1483
|
maxTokens: 65536,
|
|
1275
1484
|
costPer1MTokens: 12
|
|
1276
1485
|
},
|
|
1277
|
-
tags: ["reasoning", "agents", "general-purpose", "vision"],
|
|
1486
|
+
tags: ["reasoning", "agents", "general-purpose", "vision", "coding"],
|
|
1278
1487
|
lifecycle: "preview",
|
|
1279
1488
|
capabilities: {
|
|
1280
|
-
supportsImages: true
|
|
1489
|
+
supportsImages: true,
|
|
1490
|
+
supportsAudio: true,
|
|
1491
|
+
supportsTranscription: false,
|
|
1492
|
+
supportsSearch: true
|
|
1281
1493
|
},
|
|
1282
1494
|
aliases: ["gemini-3.1-pro-preview"]
|
|
1283
1495
|
},
|
|
@@ -1296,10 +1508,56 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1296
1508
|
tags: ["reasoning", "agents", "general-purpose", "vision"],
|
|
1297
1509
|
lifecycle: "preview",
|
|
1298
1510
|
capabilities: {
|
|
1299
|
-
supportsImages: true
|
|
1511
|
+
supportsImages: true,
|
|
1512
|
+
supportsAudio: true,
|
|
1513
|
+
supportsTranscription: false,
|
|
1514
|
+
supportsSearch: true
|
|
1300
1515
|
},
|
|
1301
1516
|
aliases: ["gemini-3-flash-preview"]
|
|
1302
1517
|
},
|
|
1518
|
+
"google-ai:gemini-3.1-flash-lite": {
|
|
1519
|
+
id: "google-ai:gemini-3.1-flash-lite",
|
|
1520
|
+
name: "Gemini 3.1 Flash-Lite",
|
|
1521
|
+
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.",
|
|
1522
|
+
input: {
|
|
1523
|
+
maxTokens: 1048576,
|
|
1524
|
+
costPer1MTokens: 0.25
|
|
1525
|
+
},
|
|
1526
|
+
output: {
|
|
1527
|
+
maxTokens: 65536,
|
|
1528
|
+
costPer1MTokens: 1.5
|
|
1529
|
+
},
|
|
1530
|
+
tags: ["low-cost", "general-purpose", "vision"],
|
|
1531
|
+
lifecycle: "preview",
|
|
1532
|
+
capabilities: {
|
|
1533
|
+
supportsImages: true,
|
|
1534
|
+
supportsAudio: true,
|
|
1535
|
+
supportsTranscription: false,
|
|
1536
|
+
supportsSearch: true
|
|
1537
|
+
},
|
|
1538
|
+
aliases: ["gemini-3.1-flash-lite-preview"]
|
|
1539
|
+
},
|
|
1540
|
+
"google-ai:gemini-2.5-pro": {
|
|
1541
|
+
id: "google-ai:gemini-2.5-pro",
|
|
1542
|
+
name: "Gemini 2.5 Pro",
|
|
1543
|
+
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.`,
|
|
1544
|
+
input: {
|
|
1545
|
+
maxTokens: 2e5,
|
|
1546
|
+
costPer1MTokens: 1.25
|
|
1547
|
+
},
|
|
1548
|
+
output: {
|
|
1549
|
+
maxTokens: 65536,
|
|
1550
|
+
costPer1MTokens: 10
|
|
1551
|
+
},
|
|
1552
|
+
tags: ["recommended", "reasoning", "agents", "general-purpose", "vision", "coding"],
|
|
1553
|
+
lifecycle: "production",
|
|
1554
|
+
capabilities: {
|
|
1555
|
+
supportsImages: true,
|
|
1556
|
+
supportsAudio: true,
|
|
1557
|
+
supportsTranscription: false,
|
|
1558
|
+
supportsSearch: true
|
|
1559
|
+
}
|
|
1560
|
+
},
|
|
1303
1561
|
"google-ai:gemini-2.5-flash": {
|
|
1304
1562
|
id: "google-ai:gemini-2.5-flash",
|
|
1305
1563
|
name: "Gemini 2.5 Flash",
|
|
@@ -1315,25 +1573,31 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1315
1573
|
tags: ["recommended", "reasoning", "agents", "general-purpose", "vision"],
|
|
1316
1574
|
lifecycle: "production",
|
|
1317
1575
|
capabilities: {
|
|
1318
|
-
supportsImages: true
|
|
1576
|
+
supportsImages: true,
|
|
1577
|
+
supportsAudio: true,
|
|
1578
|
+
supportsTranscription: false,
|
|
1579
|
+
supportsSearch: true
|
|
1319
1580
|
}
|
|
1320
1581
|
},
|
|
1321
|
-
"google-ai:gemini-2.5-
|
|
1322
|
-
id: "google-ai:gemini-2.5-
|
|
1323
|
-
name: "Gemini 2.5
|
|
1324
|
-
description:
|
|
1582
|
+
"google-ai:gemini-2.5-flash-lite": {
|
|
1583
|
+
id: "google-ai:gemini-2.5-flash-lite",
|
|
1584
|
+
name: "Gemini 2.5 Flash-Lite",
|
|
1585
|
+
description: "Lightweight, cost-efficient Gemini model optimized for high-volume, low-latency tasks. Successor to Gemini 2.0 Flash with improved capabilities.",
|
|
1325
1586
|
input: {
|
|
1326
|
-
maxTokens:
|
|
1327
|
-
costPer1MTokens: 1
|
|
1587
|
+
maxTokens: 1048576,
|
|
1588
|
+
costPer1MTokens: 0.1
|
|
1328
1589
|
},
|
|
1329
1590
|
output: {
|
|
1330
1591
|
maxTokens: 65536,
|
|
1331
|
-
costPer1MTokens:
|
|
1592
|
+
costPer1MTokens: 0.4
|
|
1332
1593
|
},
|
|
1333
|
-
tags: ["recommended", "
|
|
1594
|
+
tags: ["recommended", "low-cost", "general-purpose", "vision"],
|
|
1334
1595
|
lifecycle: "production",
|
|
1335
1596
|
capabilities: {
|
|
1336
|
-
supportsImages: true
|
|
1597
|
+
supportsImages: true,
|
|
1598
|
+
supportsAudio: true,
|
|
1599
|
+
supportsTranscription: false,
|
|
1600
|
+
supportsSearch: true
|
|
1337
1601
|
}
|
|
1338
1602
|
},
|
|
1339
1603
|
"google-ai:gemini-2.0-flash": {
|
|
@@ -1349,12 +1613,37 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1349
1613
|
costPer1MTokens: 0.4
|
|
1350
1614
|
},
|
|
1351
1615
|
tags: ["low-cost", "general-purpose", "vision"],
|
|
1352
|
-
lifecycle: "
|
|
1616
|
+
lifecycle: "deprecated",
|
|
1353
1617
|
capabilities: {
|
|
1354
|
-
supportsImages: true
|
|
1618
|
+
supportsImages: true,
|
|
1619
|
+
supportsAudio: true,
|
|
1620
|
+
supportsTranscription: false,
|
|
1621
|
+
supportsSearch: false
|
|
1355
1622
|
},
|
|
1356
1623
|
aliases: ["models/gemini-2.0-flash"]
|
|
1357
1624
|
},
|
|
1625
|
+
"google-ai:gemini-3-pro": {
|
|
1626
|
+
id: "google-ai:gemini-3-pro",
|
|
1627
|
+
name: "Gemini 3 Pro (Shut Down)",
|
|
1628
|
+
description: "Gemini 3 Pro Preview was shut down on March 9, 2026. Use Gemini 3.1 Pro instead.",
|
|
1629
|
+
input: {
|
|
1630
|
+
maxTokens: 1048576,
|
|
1631
|
+
costPer1MTokens: 2
|
|
1632
|
+
},
|
|
1633
|
+
output: {
|
|
1634
|
+
maxTokens: 65536,
|
|
1635
|
+
costPer1MTokens: 12
|
|
1636
|
+
},
|
|
1637
|
+
tags: ["reasoning", "agents", "general-purpose", "vision"],
|
|
1638
|
+
lifecycle: "discontinued",
|
|
1639
|
+
capabilities: {
|
|
1640
|
+
supportsImages: true,
|
|
1641
|
+
supportsAudio: true,
|
|
1642
|
+
supportsTranscription: false,
|
|
1643
|
+
supportsSearch: true
|
|
1644
|
+
},
|
|
1645
|
+
aliases: ["gemini-3-pro-preview"]
|
|
1646
|
+
},
|
|
1358
1647
|
"cerebras:gpt-oss-120b": {
|
|
1359
1648
|
id: "cerebras:gpt-oss-120b",
|
|
1360
1649
|
name: "GPT-OSS 120B (Preview)",
|
|
@@ -1370,7 +1659,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1370
1659
|
tags: ["preview", "general-purpose", "reasoning"],
|
|
1371
1660
|
lifecycle: "production",
|
|
1372
1661
|
capabilities: {
|
|
1373
|
-
supportsImages: false
|
|
1662
|
+
supportsImages: false,
|
|
1663
|
+
supportsAudio: false,
|
|
1664
|
+
supportsTranscription: false,
|
|
1665
|
+
supportsSearch: false
|
|
1374
1666
|
}
|
|
1375
1667
|
},
|
|
1376
1668
|
"cerebras:qwen-3-32b": {
|
|
@@ -1386,9 +1678,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1386
1678
|
costPer1MTokens: 0.8
|
|
1387
1679
|
},
|
|
1388
1680
|
tags: ["general-purpose", "reasoning"],
|
|
1389
|
-
lifecycle: "
|
|
1681
|
+
lifecycle: "discontinued",
|
|
1390
1682
|
capabilities: {
|
|
1391
|
-
supportsImages: false
|
|
1683
|
+
supportsImages: false,
|
|
1684
|
+
supportsAudio: false,
|
|
1685
|
+
supportsTranscription: false,
|
|
1686
|
+
supportsSearch: false
|
|
1392
1687
|
}
|
|
1393
1688
|
},
|
|
1394
1689
|
"cerebras:llama-4-scout-17b-16e-instruct": {
|
|
@@ -1404,9 +1699,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1404
1699
|
costPer1MTokens: 0.85
|
|
1405
1700
|
},
|
|
1406
1701
|
tags: ["general-purpose", "vision"],
|
|
1407
|
-
lifecycle: "
|
|
1702
|
+
lifecycle: "discontinued",
|
|
1408
1703
|
capabilities: {
|
|
1409
|
-
supportsImages: true
|
|
1704
|
+
supportsImages: true,
|
|
1705
|
+
supportsAudio: false,
|
|
1706
|
+
supportsTranscription: false,
|
|
1707
|
+
supportsSearch: false
|
|
1410
1708
|
}
|
|
1411
1709
|
},
|
|
1412
1710
|
"cerebras:llama3.1-8b": {
|
|
@@ -1424,7 +1722,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1424
1722
|
tags: ["low-cost", "general-purpose"],
|
|
1425
1723
|
lifecycle: "production",
|
|
1426
1724
|
capabilities: {
|
|
1427
|
-
supportsImages: false
|
|
1725
|
+
supportsImages: false,
|
|
1726
|
+
supportsAudio: false,
|
|
1727
|
+
supportsTranscription: false,
|
|
1728
|
+
supportsSearch: false
|
|
1428
1729
|
}
|
|
1429
1730
|
},
|
|
1430
1731
|
"cerebras:llama3.3-70b": {
|
|
@@ -1440,9 +1741,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1440
1741
|
costPer1MTokens: 1.2
|
|
1441
1742
|
},
|
|
1442
1743
|
tags: ["general-purpose"],
|
|
1443
|
-
lifecycle: "
|
|
1744
|
+
lifecycle: "discontinued",
|
|
1444
1745
|
capabilities: {
|
|
1445
|
-
supportsImages: false
|
|
1746
|
+
supportsImages: false,
|
|
1747
|
+
supportsAudio: false,
|
|
1748
|
+
supportsTranscription: false,
|
|
1749
|
+
supportsSearch: false
|
|
1446
1750
|
}
|
|
1447
1751
|
},
|
|
1448
1752
|
"groq:gpt-oss-20b": {
|
|
@@ -1451,16 +1755,19 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1451
1755
|
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.",
|
|
1452
1756
|
input: {
|
|
1453
1757
|
maxTokens: 131e3,
|
|
1454
|
-
costPer1MTokens: 0.
|
|
1758
|
+
costPer1MTokens: 0.075
|
|
1455
1759
|
},
|
|
1456
1760
|
output: {
|
|
1457
1761
|
maxTokens: 32e3,
|
|
1458
|
-
costPer1MTokens: 0.
|
|
1762
|
+
costPer1MTokens: 0.3
|
|
1459
1763
|
},
|
|
1460
1764
|
tags: ["preview", "general-purpose", "reasoning", "low-cost"],
|
|
1461
1765
|
lifecycle: "production",
|
|
1462
1766
|
capabilities: {
|
|
1463
|
-
supportsImages: false
|
|
1767
|
+
supportsImages: false,
|
|
1768
|
+
supportsAudio: false,
|
|
1769
|
+
supportsTranscription: false,
|
|
1770
|
+
supportsSearch: false
|
|
1464
1771
|
},
|
|
1465
1772
|
aliases: ["openai/gpt-oss-20b"]
|
|
1466
1773
|
},
|
|
@@ -1479,7 +1786,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1479
1786
|
tags: ["preview", "general-purpose", "reasoning"],
|
|
1480
1787
|
lifecycle: "production",
|
|
1481
1788
|
capabilities: {
|
|
1482
|
-
supportsImages: false
|
|
1789
|
+
supportsImages: false,
|
|
1790
|
+
supportsAudio: false,
|
|
1791
|
+
supportsTranscription: false,
|
|
1792
|
+
supportsSearch: false
|
|
1483
1793
|
},
|
|
1484
1794
|
aliases: ["openai/gpt-oss-120b"]
|
|
1485
1795
|
},
|
|
@@ -1498,7 +1808,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1498
1808
|
tags: ["general-purpose", "reasoning", "preview"],
|
|
1499
1809
|
lifecycle: "discontinued",
|
|
1500
1810
|
capabilities: {
|
|
1501
|
-
supportsImages: false
|
|
1811
|
+
supportsImages: false,
|
|
1812
|
+
supportsAudio: false,
|
|
1813
|
+
supportsTranscription: false,
|
|
1814
|
+
supportsSearch: false
|
|
1502
1815
|
}
|
|
1503
1816
|
},
|
|
1504
1817
|
"groq:llama-3.3-70b-versatile": {
|
|
@@ -1516,7 +1829,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1516
1829
|
tags: ["recommended", "general-purpose", "coding"],
|
|
1517
1830
|
lifecycle: "production",
|
|
1518
1831
|
capabilities: {
|
|
1519
|
-
supportsImages: false
|
|
1832
|
+
supportsImages: false,
|
|
1833
|
+
supportsAudio: false,
|
|
1834
|
+
supportsTranscription: false,
|
|
1835
|
+
supportsSearch: false
|
|
1520
1836
|
}
|
|
1521
1837
|
},
|
|
1522
1838
|
"groq:llama-3.2-1b-preview": {
|
|
@@ -1534,7 +1850,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1534
1850
|
tags: ["low-cost", "deprecated"],
|
|
1535
1851
|
lifecycle: "discontinued",
|
|
1536
1852
|
capabilities: {
|
|
1537
|
-
supportsImages: false
|
|
1853
|
+
supportsImages: false,
|
|
1854
|
+
supportsAudio: false,
|
|
1855
|
+
supportsTranscription: false,
|
|
1856
|
+
supportsSearch: false
|
|
1538
1857
|
}
|
|
1539
1858
|
},
|
|
1540
1859
|
"groq:llama-3.2-3b-preview": {
|
|
@@ -1552,7 +1871,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1552
1871
|
tags: ["low-cost", "general-purpose", "deprecated"],
|
|
1553
1872
|
lifecycle: "discontinued",
|
|
1554
1873
|
capabilities: {
|
|
1555
|
-
supportsImages: false
|
|
1874
|
+
supportsImages: false,
|
|
1875
|
+
supportsAudio: false,
|
|
1876
|
+
supportsTranscription: false,
|
|
1877
|
+
supportsSearch: false
|
|
1556
1878
|
}
|
|
1557
1879
|
},
|
|
1558
1880
|
"groq:llama-3.2-11b-vision-preview": {
|
|
@@ -1570,7 +1892,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1570
1892
|
tags: ["low-cost", "vision", "general-purpose", "deprecated"],
|
|
1571
1893
|
lifecycle: "discontinued",
|
|
1572
1894
|
capabilities: {
|
|
1573
|
-
supportsImages: true
|
|
1895
|
+
supportsImages: true,
|
|
1896
|
+
supportsAudio: false,
|
|
1897
|
+
supportsTranscription: false,
|
|
1898
|
+
supportsSearch: false
|
|
1574
1899
|
}
|
|
1575
1900
|
},
|
|
1576
1901
|
"groq:llama-3.2-90b-vision-preview": {
|
|
@@ -1588,7 +1913,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1588
1913
|
tags: ["vision", "general-purpose", "deprecated"],
|
|
1589
1914
|
lifecycle: "discontinued",
|
|
1590
1915
|
capabilities: {
|
|
1591
|
-
supportsImages: true
|
|
1916
|
+
supportsImages: true,
|
|
1917
|
+
supportsAudio: false,
|
|
1918
|
+
supportsTranscription: false,
|
|
1919
|
+
supportsSearch: false
|
|
1592
1920
|
}
|
|
1593
1921
|
},
|
|
1594
1922
|
"groq:llama-3.1-8b-instant": {
|
|
@@ -1606,7 +1934,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1606
1934
|
tags: ["low-cost", "general-purpose"],
|
|
1607
1935
|
lifecycle: "production",
|
|
1608
1936
|
capabilities: {
|
|
1609
|
-
supportsImages: false
|
|
1937
|
+
supportsImages: false,
|
|
1938
|
+
supportsAudio: false,
|
|
1939
|
+
supportsTranscription: false,
|
|
1940
|
+
supportsSearch: false
|
|
1610
1941
|
}
|
|
1611
1942
|
},
|
|
1612
1943
|
"groq:llama3-8b-8192": {
|
|
@@ -1624,7 +1955,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1624
1955
|
tags: ["low-cost", "general-purpose", "deprecated"],
|
|
1625
1956
|
lifecycle: "discontinued",
|
|
1626
1957
|
capabilities: {
|
|
1627
|
-
supportsImages: false
|
|
1958
|
+
supportsImages: false,
|
|
1959
|
+
supportsAudio: false,
|
|
1960
|
+
supportsTranscription: false,
|
|
1961
|
+
supportsSearch: false
|
|
1628
1962
|
}
|
|
1629
1963
|
},
|
|
1630
1964
|
"groq:llama3-70b-8192": {
|
|
@@ -1642,7 +1976,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1642
1976
|
tags: ["general-purpose", "deprecated"],
|
|
1643
1977
|
lifecycle: "discontinued",
|
|
1644
1978
|
capabilities: {
|
|
1645
|
-
supportsImages: false
|
|
1979
|
+
supportsImages: false,
|
|
1980
|
+
supportsAudio: false,
|
|
1981
|
+
supportsTranscription: false,
|
|
1982
|
+
supportsSearch: false
|
|
1646
1983
|
}
|
|
1647
1984
|
},
|
|
1648
1985
|
"groq:gemma2-9b-it": {
|
|
@@ -1660,7 +1997,160 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1660
1997
|
tags: ["low-cost", "general-purpose"],
|
|
1661
1998
|
lifecycle: "discontinued",
|
|
1662
1999
|
capabilities: {
|
|
1663
|
-
supportsImages: false
|
|
2000
|
+
supportsImages: false,
|
|
2001
|
+
supportsAudio: false,
|
|
2002
|
+
supportsTranscription: false,
|
|
2003
|
+
supportsSearch: false
|
|
2004
|
+
}
|
|
2005
|
+
},
|
|
2006
|
+
"groq:whisper-large-v3": {
|
|
2007
|
+
id: "groq:whisper-large-v3",
|
|
2008
|
+
name: "Whisper V3",
|
|
2009
|
+
description: "Whisper Large V3 on Groq \u2014 fast, accurate multilingual speech recognition.",
|
|
2010
|
+
input: {
|
|
2011
|
+
maxTokens: 0,
|
|
2012
|
+
costPer1MTokens: 0,
|
|
2013
|
+
costPerMinute: 185e-5
|
|
2014
|
+
},
|
|
2015
|
+
output: {
|
|
2016
|
+
maxTokens: 0,
|
|
2017
|
+
costPer1MTokens: 0
|
|
2018
|
+
},
|
|
2019
|
+
tags: ["general-purpose", "speech-to-text"],
|
|
2020
|
+
lifecycle: "production",
|
|
2021
|
+
capabilities: {
|
|
2022
|
+
supportsImages: false,
|
|
2023
|
+
supportsAudio: false,
|
|
2024
|
+
supportsTranscription: true,
|
|
2025
|
+
supportsSearch: false
|
|
2026
|
+
}
|
|
2027
|
+
},
|
|
2028
|
+
"groq:whisper-large-v3-turbo": {
|
|
2029
|
+
id: "groq:whisper-large-v3-turbo",
|
|
2030
|
+
name: "Whisper V3 Turbo",
|
|
2031
|
+
description: "Whisper Large V3 Turbo on Groq \u2014 optimized for speed with near-identical accuracy to V3.",
|
|
2032
|
+
input: {
|
|
2033
|
+
maxTokens: 0,
|
|
2034
|
+
costPer1MTokens: 0,
|
|
2035
|
+
costPerMinute: 667e-6
|
|
2036
|
+
},
|
|
2037
|
+
output: {
|
|
2038
|
+
maxTokens: 0,
|
|
2039
|
+
costPer1MTokens: 0
|
|
2040
|
+
},
|
|
2041
|
+
tags: ["low-cost", "general-purpose", "speech-to-text"],
|
|
2042
|
+
lifecycle: "production",
|
|
2043
|
+
capabilities: {
|
|
2044
|
+
supportsImages: false,
|
|
2045
|
+
supportsAudio: false,
|
|
2046
|
+
supportsTranscription: true,
|
|
2047
|
+
supportsSearch: false
|
|
2048
|
+
}
|
|
2049
|
+
},
|
|
2050
|
+
"groq:distil-whisper-large-v3-en": {
|
|
2051
|
+
id: "groq:distil-whisper-large-v3-en",
|
|
2052
|
+
name: "Distil Whisper V3 (English)",
|
|
2053
|
+
description: "Distilled Whisper Large V3 on Groq \u2014 decommissioned, replaced by whisper-large-v3-turbo.",
|
|
2054
|
+
input: {
|
|
2055
|
+
maxTokens: 0,
|
|
2056
|
+
costPer1MTokens: 0,
|
|
2057
|
+
costPerMinute: 333e-6
|
|
2058
|
+
},
|
|
2059
|
+
output: {
|
|
2060
|
+
maxTokens: 0,
|
|
2061
|
+
costPer1MTokens: 0
|
|
2062
|
+
},
|
|
2063
|
+
tags: ["low-cost", "speech-to-text"],
|
|
2064
|
+
lifecycle: "discontinued",
|
|
2065
|
+
capabilities: {
|
|
2066
|
+
supportsImages: false,
|
|
2067
|
+
supportsAudio: false,
|
|
2068
|
+
supportsTranscription: true,
|
|
2069
|
+
supportsSearch: false
|
|
2070
|
+
}
|
|
2071
|
+
},
|
|
2072
|
+
"xai:grok-4-1-fast-reasoning": {
|
|
2073
|
+
id: "xai:grok-4-1-fast-reasoning",
|
|
2074
|
+
name: "Grok 4.1 Fast (Reasoning)",
|
|
2075
|
+
description: "Latest fast Grok model with reasoning capabilities and a massive 2M context window. Extremely cost-effective for a frontier-class model.",
|
|
2076
|
+
input: {
|
|
2077
|
+
maxTokens: 2e6,
|
|
2078
|
+
costPer1MTokens: 0.2
|
|
2079
|
+
},
|
|
2080
|
+
output: {
|
|
2081
|
+
maxTokens: 128e3,
|
|
2082
|
+
costPer1MTokens: 0.5
|
|
2083
|
+
},
|
|
2084
|
+
tags: ["recommended", "reasoning", "general-purpose", "vision", "low-cost"],
|
|
2085
|
+
lifecycle: "production",
|
|
2086
|
+
capabilities: {
|
|
2087
|
+
supportsImages: true,
|
|
2088
|
+
supportsAudio: false,
|
|
2089
|
+
supportsTranscription: false,
|
|
2090
|
+
supportsSearch: false
|
|
2091
|
+
}
|
|
2092
|
+
},
|
|
2093
|
+
"xai:grok-4-1-fast-non-reasoning": {
|
|
2094
|
+
id: "xai:grok-4-1-fast-non-reasoning",
|
|
2095
|
+
name: "Grok 4.1 Fast (Non-Reasoning)",
|
|
2096
|
+
description: "Latest fast Grok model for non-reasoning tasks with a massive 2M context window. Extremely cost-effective for a frontier-class model.",
|
|
2097
|
+
input: {
|
|
2098
|
+
maxTokens: 2e6,
|
|
2099
|
+
costPer1MTokens: 0.2
|
|
2100
|
+
},
|
|
2101
|
+
output: {
|
|
2102
|
+
maxTokens: 128e3,
|
|
2103
|
+
costPer1MTokens: 0.5
|
|
2104
|
+
},
|
|
2105
|
+
tags: ["recommended", "general-purpose", "vision", "low-cost"],
|
|
2106
|
+
lifecycle: "production",
|
|
2107
|
+
capabilities: {
|
|
2108
|
+
supportsImages: true,
|
|
2109
|
+
supportsAudio: false,
|
|
2110
|
+
supportsTranscription: false,
|
|
2111
|
+
supportsSearch: false
|
|
2112
|
+
}
|
|
2113
|
+
},
|
|
2114
|
+
"xai:grok-4.20-0309-reasoning": {
|
|
2115
|
+
id: "xai:grok-4.20-0309-reasoning",
|
|
2116
|
+
name: "Grok 4.20 (Reasoning)",
|
|
2117
|
+
description: "xAI flagship model with deep reasoning capabilities and 2M context window.",
|
|
2118
|
+
input: {
|
|
2119
|
+
maxTokens: 2e6,
|
|
2120
|
+
costPer1MTokens: 2
|
|
2121
|
+
},
|
|
2122
|
+
output: {
|
|
2123
|
+
maxTokens: 128e3,
|
|
2124
|
+
costPer1MTokens: 6
|
|
2125
|
+
},
|
|
2126
|
+
tags: ["reasoning", "general-purpose", "vision"],
|
|
2127
|
+
lifecycle: "production",
|
|
2128
|
+
capabilities: {
|
|
2129
|
+
supportsImages: true,
|
|
2130
|
+
supportsAudio: false,
|
|
2131
|
+
supportsTranscription: false,
|
|
2132
|
+
supportsSearch: false
|
|
2133
|
+
}
|
|
2134
|
+
},
|
|
2135
|
+
"xai:grok-4.20-0309-non-reasoning": {
|
|
2136
|
+
id: "xai:grok-4.20-0309-non-reasoning",
|
|
2137
|
+
name: "Grok 4.20 (Non-Reasoning)",
|
|
2138
|
+
description: "xAI flagship model for non-reasoning tasks with 2M context window.",
|
|
2139
|
+
input: {
|
|
2140
|
+
maxTokens: 2e6,
|
|
2141
|
+
costPer1MTokens: 2
|
|
2142
|
+
},
|
|
2143
|
+
output: {
|
|
2144
|
+
maxTokens: 128e3,
|
|
2145
|
+
costPer1MTokens: 6
|
|
2146
|
+
},
|
|
2147
|
+
tags: ["general-purpose", "vision"],
|
|
2148
|
+
lifecycle: "production",
|
|
2149
|
+
capabilities: {
|
|
2150
|
+
supportsImages: true,
|
|
2151
|
+
supportsAudio: false,
|
|
2152
|
+
supportsTranscription: false,
|
|
2153
|
+
supportsSearch: false
|
|
1664
2154
|
}
|
|
1665
2155
|
},
|
|
1666
2156
|
"xai:grok-code-fast-1": {
|
|
@@ -1678,7 +2168,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1678
2168
|
tags: ["coding", "general-purpose", "low-cost"],
|
|
1679
2169
|
lifecycle: "production",
|
|
1680
2170
|
capabilities: {
|
|
1681
|
-
supportsImages: false
|
|
2171
|
+
supportsImages: false,
|
|
2172
|
+
supportsAudio: false,
|
|
2173
|
+
supportsTranscription: false,
|
|
2174
|
+
supportsSearch: false
|
|
1682
2175
|
}
|
|
1683
2176
|
},
|
|
1684
2177
|
"xai:grok-4-fast-reasoning": {
|
|
@@ -1693,10 +2186,13 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1693
2186
|
maxTokens: 128e3,
|
|
1694
2187
|
costPer1MTokens: 0.5
|
|
1695
2188
|
},
|
|
1696
|
-
tags: ["reasoning", "
|
|
2189
|
+
tags: ["reasoning", "general-purpose"],
|
|
1697
2190
|
lifecycle: "production",
|
|
1698
2191
|
capabilities: {
|
|
1699
|
-
supportsImages: true
|
|
2192
|
+
supportsImages: true,
|
|
2193
|
+
supportsAudio: false,
|
|
2194
|
+
supportsTranscription: false,
|
|
2195
|
+
supportsSearch: false
|
|
1700
2196
|
}
|
|
1701
2197
|
},
|
|
1702
2198
|
"xai:grok-4-fast-non-reasoning": {
|
|
@@ -1711,10 +2207,13 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1711
2207
|
maxTokens: 128e3,
|
|
1712
2208
|
costPer1MTokens: 0.5
|
|
1713
2209
|
},
|
|
1714
|
-
tags: ["low-cost", "
|
|
2210
|
+
tags: ["low-cost", "general-purpose"],
|
|
1715
2211
|
lifecycle: "production",
|
|
1716
2212
|
capabilities: {
|
|
1717
|
-
supportsImages: true
|
|
2213
|
+
supportsImages: true,
|
|
2214
|
+
supportsAudio: false,
|
|
2215
|
+
supportsTranscription: false,
|
|
2216
|
+
supportsSearch: false
|
|
1718
2217
|
}
|
|
1719
2218
|
},
|
|
1720
2219
|
"xai:grok-4-0709": {
|
|
@@ -1732,7 +2231,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1732
2231
|
tags: ["reasoning", "general-purpose"],
|
|
1733
2232
|
lifecycle: "production",
|
|
1734
2233
|
capabilities: {
|
|
1735
|
-
supportsImages: true
|
|
2234
|
+
supportsImages: true,
|
|
2235
|
+
supportsAudio: false,
|
|
2236
|
+
supportsTranscription: false,
|
|
2237
|
+
supportsSearch: false
|
|
1736
2238
|
}
|
|
1737
2239
|
},
|
|
1738
2240
|
"xai:grok-3-mini": {
|
|
@@ -1750,7 +2252,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1750
2252
|
tags: ["low-cost", "general-purpose"],
|
|
1751
2253
|
lifecycle: "production",
|
|
1752
2254
|
capabilities: {
|
|
1753
|
-
supportsImages: false
|
|
2255
|
+
supportsImages: false,
|
|
2256
|
+
supportsAudio: false,
|
|
2257
|
+
supportsTranscription: false,
|
|
2258
|
+
supportsSearch: false
|
|
1754
2259
|
}
|
|
1755
2260
|
},
|
|
1756
2261
|
"xai:grok-3": {
|
|
@@ -1768,7 +2273,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1768
2273
|
tags: ["general-purpose"],
|
|
1769
2274
|
lifecycle: "production",
|
|
1770
2275
|
capabilities: {
|
|
1771
|
-
supportsImages: true
|
|
2276
|
+
supportsImages: true,
|
|
2277
|
+
supportsAudio: false,
|
|
2278
|
+
supportsTranscription: false,
|
|
2279
|
+
supportsSearch: false
|
|
1772
2280
|
}
|
|
1773
2281
|
},
|
|
1774
2282
|
"openrouter:gpt-oss-120b": {
|
|
@@ -1786,9 +2294,34 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1786
2294
|
tags: ["preview", "general-purpose", "reasoning"],
|
|
1787
2295
|
lifecycle: "production",
|
|
1788
2296
|
capabilities: {
|
|
1789
|
-
supportsImages: false
|
|
2297
|
+
supportsImages: false,
|
|
2298
|
+
supportsAudio: false,
|
|
2299
|
+
supportsTranscription: false,
|
|
2300
|
+
supportsSearch: false
|
|
1790
2301
|
}
|
|
1791
2302
|
},
|
|
2303
|
+
"fireworks-ai:qwen3-8b": {
|
|
2304
|
+
id: "fireworks-ai:qwen3-8b",
|
|
2305
|
+
name: "Qwen3 8B",
|
|
2306
|
+
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.",
|
|
2307
|
+
input: {
|
|
2308
|
+
maxTokens: 40960,
|
|
2309
|
+
costPer1MTokens: 0.2
|
|
2310
|
+
},
|
|
2311
|
+
output: {
|
|
2312
|
+
maxTokens: 16384,
|
|
2313
|
+
costPer1MTokens: 0.2
|
|
2314
|
+
},
|
|
2315
|
+
tags: ["low-cost", "general-purpose", "reasoning"],
|
|
2316
|
+
lifecycle: "production",
|
|
2317
|
+
capabilities: {
|
|
2318
|
+
supportsImages: false,
|
|
2319
|
+
supportsAudio: false,
|
|
2320
|
+
supportsTranscription: false,
|
|
2321
|
+
supportsSearch: false
|
|
2322
|
+
},
|
|
2323
|
+
aliases: ["accounts/fireworks/models/qwen3-8b"]
|
|
2324
|
+
},
|
|
1792
2325
|
"fireworks-ai:gpt-oss-20b": {
|
|
1793
2326
|
id: "fireworks-ai:gpt-oss-20b",
|
|
1794
2327
|
name: "GPT-OSS 20B",
|
|
@@ -1804,7 +2337,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1804
2337
|
tags: ["general-purpose", "reasoning", "low-cost"],
|
|
1805
2338
|
lifecycle: "production",
|
|
1806
2339
|
capabilities: {
|
|
1807
|
-
supportsImages: false
|
|
2340
|
+
supportsImages: false,
|
|
2341
|
+
supportsAudio: false,
|
|
2342
|
+
supportsTranscription: false,
|
|
2343
|
+
supportsSearch: false
|
|
1808
2344
|
},
|
|
1809
2345
|
aliases: ["accounts/fireworks/models/gpt-oss-20b"]
|
|
1810
2346
|
},
|
|
@@ -1823,10 +2359,57 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1823
2359
|
tags: ["general-purpose", "reasoning"],
|
|
1824
2360
|
lifecycle: "production",
|
|
1825
2361
|
capabilities: {
|
|
1826
|
-
supportsImages: false
|
|
2362
|
+
supportsImages: false,
|
|
2363
|
+
supportsAudio: false,
|
|
2364
|
+
supportsTranscription: false,
|
|
2365
|
+
supportsSearch: false
|
|
1827
2366
|
},
|
|
1828
2367
|
aliases: ["accounts/fireworks/models/gpt-oss-120b"]
|
|
1829
2368
|
},
|
|
2369
|
+
"fireworks-ai:deepseek-v3p2": {
|
|
2370
|
+
id: "fireworks-ai:deepseek-v3p2",
|
|
2371
|
+
name: "DeepSeek V3.2",
|
|
2372
|
+
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.",
|
|
2373
|
+
input: {
|
|
2374
|
+
maxTokens: 163840,
|
|
2375
|
+
costPer1MTokens: 0.56
|
|
2376
|
+
},
|
|
2377
|
+
output: {
|
|
2378
|
+
maxTokens: 16e4,
|
|
2379
|
+
costPer1MTokens: 1.68
|
|
2380
|
+
},
|
|
2381
|
+
tags: ["recommended", "reasoning", "general-purpose", "coding"],
|
|
2382
|
+
lifecycle: "production",
|
|
2383
|
+
capabilities: {
|
|
2384
|
+
supportsImages: false,
|
|
2385
|
+
supportsAudio: false,
|
|
2386
|
+
supportsTranscription: false,
|
|
2387
|
+
supportsSearch: false
|
|
2388
|
+
},
|
|
2389
|
+
aliases: ["accounts/fireworks/models/deepseek-v3p2"]
|
|
2390
|
+
},
|
|
2391
|
+
"fireworks-ai:deepseek-v3p1": {
|
|
2392
|
+
id: "fireworks-ai:deepseek-v3p1",
|
|
2393
|
+
name: "DeepSeek V3.1",
|
|
2394
|
+
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.",
|
|
2395
|
+
input: {
|
|
2396
|
+
maxTokens: 163840,
|
|
2397
|
+
costPer1MTokens: 0.56
|
|
2398
|
+
},
|
|
2399
|
+
output: {
|
|
2400
|
+
maxTokens: 163840,
|
|
2401
|
+
costPer1MTokens: 1.68
|
|
2402
|
+
},
|
|
2403
|
+
tags: ["reasoning", "general-purpose", "coding"],
|
|
2404
|
+
lifecycle: "production",
|
|
2405
|
+
capabilities: {
|
|
2406
|
+
supportsImages: false,
|
|
2407
|
+
supportsAudio: false,
|
|
2408
|
+
supportsTranscription: false,
|
|
2409
|
+
supportsSearch: false
|
|
2410
|
+
},
|
|
2411
|
+
aliases: ["accounts/fireworks/models/deepseek-v3p1"]
|
|
2412
|
+
},
|
|
1830
2413
|
"fireworks-ai:deepseek-r1-0528": {
|
|
1831
2414
|
id: "fireworks-ai:deepseek-r1-0528",
|
|
1832
2415
|
name: "DeepSeek R1 0528",
|
|
@@ -1840,9 +2423,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1840
2423
|
costPer1MTokens: 8
|
|
1841
2424
|
},
|
|
1842
2425
|
tags: ["recommended", "reasoning", "general-purpose", "coding"],
|
|
1843
|
-
lifecycle: "
|
|
2426
|
+
lifecycle: "discontinued",
|
|
1844
2427
|
capabilities: {
|
|
1845
|
-
supportsImages: false
|
|
2428
|
+
supportsImages: false,
|
|
2429
|
+
supportsAudio: false,
|
|
2430
|
+
supportsTranscription: false,
|
|
2431
|
+
supportsSearch: false
|
|
1846
2432
|
},
|
|
1847
2433
|
aliases: ["accounts/fireworks/models/deepseek-r1-0528"]
|
|
1848
2434
|
},
|
|
@@ -1859,9 +2445,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1859
2445
|
costPer1MTokens: 0.9
|
|
1860
2446
|
},
|
|
1861
2447
|
tags: ["recommended", "general-purpose"],
|
|
1862
|
-
lifecycle: "
|
|
2448
|
+
lifecycle: "discontinued",
|
|
1863
2449
|
capabilities: {
|
|
1864
|
-
supportsImages: false
|
|
2450
|
+
supportsImages: false,
|
|
2451
|
+
supportsAudio: false,
|
|
2452
|
+
supportsTranscription: false,
|
|
2453
|
+
supportsSearch: false
|
|
1865
2454
|
},
|
|
1866
2455
|
aliases: ["accounts/fireworks/models/deepseek-v3-0324"]
|
|
1867
2456
|
},
|
|
@@ -1878,9 +2467,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1878
2467
|
costPer1MTokens: 0.88
|
|
1879
2468
|
},
|
|
1880
2469
|
tags: ["general-purpose", "vision"],
|
|
1881
|
-
lifecycle: "
|
|
2470
|
+
lifecycle: "discontinued",
|
|
1882
2471
|
capabilities: {
|
|
1883
|
-
supportsImages: true
|
|
2472
|
+
supportsImages: true,
|
|
2473
|
+
supportsAudio: false,
|
|
2474
|
+
supportsTranscription: false,
|
|
2475
|
+
supportsSearch: false
|
|
1884
2476
|
},
|
|
1885
2477
|
aliases: ["accounts/fireworks/models/llama4-maverick-instruct-basic"]
|
|
1886
2478
|
},
|
|
@@ -1897,9 +2489,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1897
2489
|
costPer1MTokens: 0.6
|
|
1898
2490
|
},
|
|
1899
2491
|
tags: ["general-purpose", "vision"],
|
|
1900
|
-
lifecycle: "
|
|
2492
|
+
lifecycle: "discontinued",
|
|
1901
2493
|
capabilities: {
|
|
1902
|
-
supportsImages: true
|
|
2494
|
+
supportsImages: true,
|
|
2495
|
+
supportsAudio: false,
|
|
2496
|
+
supportsTranscription: false,
|
|
2497
|
+
supportsSearch: false
|
|
1903
2498
|
},
|
|
1904
2499
|
aliases: ["accounts/fireworks/models/llama4-scout-instruct-basic"]
|
|
1905
2500
|
},
|
|
@@ -1918,7 +2513,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1918
2513
|
tags: ["general-purpose"],
|
|
1919
2514
|
lifecycle: "production",
|
|
1920
2515
|
capabilities: {
|
|
1921
|
-
supportsImages: false
|
|
2516
|
+
supportsImages: false,
|
|
2517
|
+
supportsAudio: false,
|
|
2518
|
+
supportsTranscription: false,
|
|
2519
|
+
supportsSearch: false
|
|
1922
2520
|
},
|
|
1923
2521
|
aliases: ["accounts/fireworks/models/llama-v3p3-70b-instruct"]
|
|
1924
2522
|
},
|
|
@@ -1937,7 +2535,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1937
2535
|
tags: ["reasoning", "general-purpose", "coding"],
|
|
1938
2536
|
lifecycle: "discontinued",
|
|
1939
2537
|
capabilities: {
|
|
1940
|
-
supportsImages: false
|
|
2538
|
+
supportsImages: false,
|
|
2539
|
+
supportsAudio: false,
|
|
2540
|
+
supportsTranscription: false,
|
|
2541
|
+
supportsSearch: false
|
|
1941
2542
|
},
|
|
1942
2543
|
aliases: ["accounts/fireworks/models/deepseek-r1"]
|
|
1943
2544
|
},
|
|
@@ -1956,7 +2557,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1956
2557
|
tags: ["reasoning", "general-purpose", "coding"],
|
|
1957
2558
|
lifecycle: "discontinued",
|
|
1958
2559
|
capabilities: {
|
|
1959
|
-
supportsImages: false
|
|
2560
|
+
supportsImages: false,
|
|
2561
|
+
supportsAudio: false,
|
|
2562
|
+
supportsTranscription: false,
|
|
2563
|
+
supportsSearch: false
|
|
1960
2564
|
},
|
|
1961
2565
|
aliases: ["accounts/fireworks/models/deepseek-r1-basic"]
|
|
1962
2566
|
},
|
|
@@ -1975,7 +2579,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1975
2579
|
tags: ["deprecated", "general-purpose"],
|
|
1976
2580
|
lifecycle: "discontinued",
|
|
1977
2581
|
capabilities: {
|
|
1978
|
-
supportsImages: false
|
|
2582
|
+
supportsImages: false,
|
|
2583
|
+
supportsAudio: false,
|
|
2584
|
+
supportsTranscription: false,
|
|
2585
|
+
supportsSearch: false
|
|
1979
2586
|
},
|
|
1980
2587
|
aliases: ["accounts/fireworks/models/deepseek-v3"]
|
|
1981
2588
|
},
|
|
@@ -1994,7 +2601,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
1994
2601
|
tags: ["deprecated", "general-purpose"],
|
|
1995
2602
|
lifecycle: "discontinued",
|
|
1996
2603
|
capabilities: {
|
|
1997
|
-
supportsImages: false
|
|
2604
|
+
supportsImages: false,
|
|
2605
|
+
supportsAudio: false,
|
|
2606
|
+
supportsTranscription: false,
|
|
2607
|
+
supportsSearch: false
|
|
1998
2608
|
},
|
|
1999
2609
|
aliases: ["accounts/fireworks/models/llama-v3p1-405b-instruct"]
|
|
2000
2610
|
},
|
|
@@ -2013,7 +2623,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2013
2623
|
tags: ["deprecated", "general-purpose"],
|
|
2014
2624
|
lifecycle: "discontinued",
|
|
2015
2625
|
capabilities: {
|
|
2016
|
-
supportsImages: false
|
|
2626
|
+
supportsImages: false,
|
|
2627
|
+
supportsAudio: false,
|
|
2628
|
+
supportsTranscription: false,
|
|
2629
|
+
supportsSearch: false
|
|
2017
2630
|
},
|
|
2018
2631
|
aliases: ["accounts/fireworks/models/llama-v3p1-70b-instruct"]
|
|
2019
2632
|
},
|
|
@@ -2030,9 +2643,12 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2030
2643
|
costPer1MTokens: 0.2
|
|
2031
2644
|
},
|
|
2032
2645
|
tags: ["low-cost", "general-purpose"],
|
|
2033
|
-
lifecycle: "
|
|
2646
|
+
lifecycle: "discontinued",
|
|
2034
2647
|
capabilities: {
|
|
2035
|
-
supportsImages: false
|
|
2648
|
+
supportsImages: false,
|
|
2649
|
+
supportsAudio: false,
|
|
2650
|
+
supportsTranscription: false,
|
|
2651
|
+
supportsSearch: false
|
|
2036
2652
|
},
|
|
2037
2653
|
aliases: ["accounts/fireworks/models/llama-v3p1-8b-instruct"]
|
|
2038
2654
|
},
|
|
@@ -2051,7 +2667,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2051
2667
|
tags: ["deprecated", "general-purpose"],
|
|
2052
2668
|
lifecycle: "discontinued",
|
|
2053
2669
|
capabilities: {
|
|
2054
|
-
supportsImages: false
|
|
2670
|
+
supportsImages: false,
|
|
2671
|
+
supportsAudio: false,
|
|
2672
|
+
supportsTranscription: false,
|
|
2673
|
+
supportsSearch: false
|
|
2055
2674
|
},
|
|
2056
2675
|
aliases: ["accounts/fireworks/models/mixtral-8x22b-instruct"]
|
|
2057
2676
|
},
|
|
@@ -2070,7 +2689,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2070
2689
|
tags: ["low-cost", "general-purpose"],
|
|
2071
2690
|
lifecycle: "discontinued",
|
|
2072
2691
|
capabilities: {
|
|
2073
|
-
supportsImages: false
|
|
2692
|
+
supportsImages: false,
|
|
2693
|
+
supportsAudio: false,
|
|
2694
|
+
supportsTranscription: false,
|
|
2695
|
+
supportsSearch: false
|
|
2074
2696
|
},
|
|
2075
2697
|
aliases: ["accounts/fireworks/models/mixtral-8x7b-instruct"]
|
|
2076
2698
|
},
|
|
@@ -2089,7 +2711,10 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2089
2711
|
tags: ["roleplay", "storytelling", "low-cost"],
|
|
2090
2712
|
lifecycle: "discontinued",
|
|
2091
2713
|
capabilities: {
|
|
2092
|
-
supportsImages: false
|
|
2714
|
+
supportsImages: false,
|
|
2715
|
+
supportsAudio: false,
|
|
2716
|
+
supportsTranscription: false,
|
|
2717
|
+
supportsSearch: false
|
|
2093
2718
|
},
|
|
2094
2719
|
aliases: ["accounts/fireworks/models/mythomax-l2-13b"]
|
|
2095
2720
|
},
|
|
@@ -2108,9 +2733,34 @@ Claude Haiku 4.5 is Anthropic's fastest and most efficient model, delivering nea
|
|
|
2108
2733
|
tags: ["deprecated", "low-cost", "general-purpose"],
|
|
2109
2734
|
lifecycle: "discontinued",
|
|
2110
2735
|
capabilities: {
|
|
2111
|
-
supportsImages: false
|
|
2736
|
+
supportsImages: false,
|
|
2737
|
+
supportsAudio: false,
|
|
2738
|
+
supportsTranscription: false,
|
|
2739
|
+
supportsSearch: false
|
|
2112
2740
|
},
|
|
2113
2741
|
aliases: ["accounts/fireworks/models/gemma2-9b-it"]
|
|
2742
|
+
},
|
|
2743
|
+
"fireworks-ai:whisper-v3": {
|
|
2744
|
+
id: "fireworks-ai:whisper-v3",
|
|
2745
|
+
name: "Whisper V3",
|
|
2746
|
+
description: "Whisper V3 on Fireworks AI \u2014 multilingual speech recognition with high accuracy.",
|
|
2747
|
+
input: {
|
|
2748
|
+
maxTokens: 0,
|
|
2749
|
+
costPer1MTokens: 0,
|
|
2750
|
+
costPerMinute: 15e-4
|
|
2751
|
+
},
|
|
2752
|
+
output: {
|
|
2753
|
+
maxTokens: 0,
|
|
2754
|
+
costPer1MTokens: 0
|
|
2755
|
+
},
|
|
2756
|
+
tags: ["general-purpose", "speech-to-text"],
|
|
2757
|
+
lifecycle: "production",
|
|
2758
|
+
capabilities: {
|
|
2759
|
+
supportsImages: false,
|
|
2760
|
+
supportsAudio: false,
|
|
2761
|
+
supportsTranscription: true,
|
|
2762
|
+
supportsSearch: false
|
|
2763
|
+
}
|
|
2114
2764
|
}
|
|
2115
2765
|
};
|
|
2116
2766
|
var defaultModel = {
|
|
@@ -2174,7 +2824,7 @@ var CognitiveBeta = class _CognitiveBeta {
|
|
|
2174
2824
|
}
|
|
2175
2825
|
async generateText(input, options = {}) {
|
|
2176
2826
|
const signal = options.signal ?? AbortSignal.timeout(this._timeout);
|
|
2177
|
-
const req = { input };
|
|
2827
|
+
const req = { type: "generateText", input };
|
|
2178
2828
|
this._events.emit("request", req);
|
|
2179
2829
|
try {
|
|
2180
2830
|
const { data } = await this._withServerRetry(
|
|
@@ -2198,9 +2848,32 @@ var CognitiveBeta = class _CognitiveBeta {
|
|
|
2198
2848
|
);
|
|
2199
2849
|
return data.models;
|
|
2200
2850
|
}
|
|
2851
|
+
async transcribeAudio(input, options = {}) {
|
|
2852
|
+
const signal = options.signal ?? AbortSignal.timeout(this._timeout);
|
|
2853
|
+
const req = { type: "transcribeAudio", input };
|
|
2854
|
+
this._events.emit("request", req);
|
|
2855
|
+
try {
|
|
2856
|
+
const { data } = await this._withServerRetry(
|
|
2857
|
+
() => this._axiosClient.post("/v2/cognitive/transcribe-audio", input, {
|
|
2858
|
+
signal,
|
|
2859
|
+
timeout: options.timeout ?? this._timeout
|
|
2860
|
+
}),
|
|
2861
|
+
options,
|
|
2862
|
+
req
|
|
2863
|
+
);
|
|
2864
|
+
if (data.error) {
|
|
2865
|
+
throw new Error(`Transcription error: ${data.error}`);
|
|
2866
|
+
}
|
|
2867
|
+
this._events.emit("response", req, data);
|
|
2868
|
+
return data;
|
|
2869
|
+
} catch (error) {
|
|
2870
|
+
this._events.emit("error", req, error);
|
|
2871
|
+
throw error;
|
|
2872
|
+
}
|
|
2873
|
+
}
|
|
2201
2874
|
async *generateTextStream(request, options = {}) {
|
|
2202
2875
|
const signal = options.signal ?? AbortSignal.timeout(this._timeout);
|
|
2203
|
-
const req = { input: request };
|
|
2876
|
+
const req = { type: "generateText", input: request };
|
|
2204
2877
|
const chunks = [];
|
|
2205
2878
|
let lastChunk;
|
|
2206
2879
|
this._events.emit("request", req);
|