ruby_llm 1.1.1 → 1.2.0
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.
- checksums.yaml +4 -4
- data/README.md +2 -2
- data/lib/ruby_llm/chat.rb +19 -6
- data/lib/ruby_llm/configuration.rb +2 -1
- data/lib/ruby_llm/models.json +416 -80
- data/lib/ruby_llm/providers/bedrock/models.rb +8 -1
- data/lib/ruby_llm/providers/openai/capabilities.rb +23 -8
- data/lib/ruby_llm/providers/openai.rb +1 -1
- data/lib/ruby_llm/version.rb +1 -1
- data/lib/ruby_llm.rb +2 -2
- data/lib/tasks/models_docs.rake +13 -1
- metadata +2 -2
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 2f06ce431337dc189e6172b0c98ed897fdba930200f3f118c39c15f4527ec135
|
|
4
|
+
data.tar.gz: 18f8ff36e7ee18cbee315e66db4b8f04619c98595a5de3b73d215bed248ca0d4
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 42f7603cfec24fa6cc59b1186d2d6a90af9e9076eb79124ac5ce09d73000fbcdb931ab90cafe21bf95b39417a52ff000ca9d02ba51c8a78877d1a1f47b70866f
|
|
7
|
+
data.tar.gz: 8513b6774ef3d745e7bbc8947f856608d13ade163ca658139c2992e923ad08d5ee00b99275cbfe762591182897b73db7bfc879fae8db7b8ce2f3ba1fea5ee235
|
data/README.md
CHANGED
|
@@ -118,7 +118,7 @@ end
|
|
|
118
118
|
## Have great conversations
|
|
119
119
|
|
|
120
120
|
```ruby
|
|
121
|
-
# Start a chat with the default model (
|
|
121
|
+
# Start a chat with the default model (gpt-4.1-nano)
|
|
122
122
|
chat = RubyLLM.chat
|
|
123
123
|
|
|
124
124
|
# Or specify what you want
|
|
@@ -169,7 +169,7 @@ class ToolCall < ApplicationRecord
|
|
|
169
169
|
end
|
|
170
170
|
|
|
171
171
|
# In a background job
|
|
172
|
-
chat = Chat.create! model_id: "gpt-
|
|
172
|
+
chat = Chat.create! model_id: "gpt-4.1-nano"
|
|
173
173
|
|
|
174
174
|
# Set personality or behavior with instructions (aka system prompts) - they're persisted too!
|
|
175
175
|
chat.with_instructions "You are a friendly Ruby expert who loves to help beginners"
|
data/lib/ruby_llm/chat.rb
CHANGED
|
@@ -8,14 +8,18 @@ module RubyLLM
|
|
|
8
8
|
# chat = RubyLLM.chat
|
|
9
9
|
# chat.ask "What's the best way to learn Ruby?"
|
|
10
10
|
# chat.ask "Can you elaborate on that?"
|
|
11
|
-
class Chat
|
|
11
|
+
class Chat # rubocop:disable Metrics/ClassLength
|
|
12
12
|
include Enumerable
|
|
13
13
|
|
|
14
14
|
attr_reader :model, :messages, :tools
|
|
15
15
|
|
|
16
|
-
def initialize(model: nil, provider: nil)
|
|
16
|
+
def initialize(model: nil, provider: nil, assume_model_exists: false) # rubocop:disable Metrics/MethodLength
|
|
17
|
+
if assume_model_exists && !provider
|
|
18
|
+
raise ArgumentError, 'Provider must be specified if assume_model_exists is true'
|
|
19
|
+
end
|
|
20
|
+
|
|
17
21
|
model_id = model || RubyLLM.config.default_model
|
|
18
|
-
with_model(model_id, provider: provider)
|
|
22
|
+
with_model(model_id, provider: provider, assume_exists: assume_model_exists)
|
|
19
23
|
@temperature = 0.7
|
|
20
24
|
@messages = []
|
|
21
25
|
@tools = {}
|
|
@@ -54,9 +58,18 @@ module RubyLLM
|
|
|
54
58
|
self
|
|
55
59
|
end
|
|
56
60
|
|
|
57
|
-
def with_model(model_id, provider: nil)
|
|
58
|
-
|
|
59
|
-
|
|
61
|
+
def with_model(model_id, provider: nil, assume_exists: false) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
|
|
62
|
+
if assume_exists
|
|
63
|
+
raise ArgumentError, 'Provider must be specified if assume_exists is true' unless provider
|
|
64
|
+
|
|
65
|
+
@provider = Provider.providers[provider.to_sym] || raise(Error, "Unknown provider: #{provider.to_sym}")
|
|
66
|
+
@model = Struct.new(:id, :provider, :supports_functions, :supports_vision).new(model_id, provider, true, true)
|
|
67
|
+
RubyLLM.logger.warn "Assuming model '#{model_id}' exists for provider '#{provider}'. " \
|
|
68
|
+
'Capabilities may not be accurately reflected.'
|
|
69
|
+
else
|
|
70
|
+
@model = Models.find model_id, provider
|
|
71
|
+
@provider = Provider.providers[@model.provider.to_sym] || raise(Error, "Unknown provider: #{@model.provider}")
|
|
72
|
+
end
|
|
60
73
|
self
|
|
61
74
|
end
|
|
62
75
|
|
|
@@ -12,6 +12,7 @@ module RubyLLM
|
|
|
12
12
|
class Configuration
|
|
13
13
|
# Provider-specific configuration
|
|
14
14
|
attr_accessor :openai_api_key,
|
|
15
|
+
:openai_api_base,
|
|
15
16
|
:anthropic_api_key,
|
|
16
17
|
:gemini_api_key,
|
|
17
18
|
:deepseek_api_key,
|
|
@@ -39,7 +40,7 @@ module RubyLLM
|
|
|
39
40
|
@retry_interval_randomness = 0.5
|
|
40
41
|
|
|
41
42
|
# Default models
|
|
42
|
-
@default_model = 'gpt-
|
|
43
|
+
@default_model = 'gpt-4.1-nano'
|
|
43
44
|
@default_embedding_model = 'text-embedding-3-small'
|
|
44
45
|
@default_image_model = 'dall-e-3'
|
|
45
46
|
end
|
data/lib/ruby_llm/models.json
CHANGED
|
@@ -4,7 +4,7 @@
|
|
|
4
4
|
"created_at": null,
|
|
5
5
|
"display_name": "Claude 3.5 Haiku",
|
|
6
6
|
"provider": "bedrock",
|
|
7
|
-
"context_window":
|
|
7
|
+
"context_window": 200000,
|
|
8
8
|
"max_tokens": 4096,
|
|
9
9
|
"type": "chat",
|
|
10
10
|
"family": "claude3_5_haiku",
|
|
@@ -34,7 +34,7 @@
|
|
|
34
34
|
"created_at": null,
|
|
35
35
|
"display_name": "Claude 3.5 Sonnet",
|
|
36
36
|
"provider": "bedrock",
|
|
37
|
-
"context_window":
|
|
37
|
+
"context_window": 200000,
|
|
38
38
|
"max_tokens": 4096,
|
|
39
39
|
"type": "chat",
|
|
40
40
|
"family": "claude3_sonnet",
|
|
@@ -47,7 +47,8 @@
|
|
|
47
47
|
"provider_name": "Anthropic",
|
|
48
48
|
"customizations_supported": [],
|
|
49
49
|
"inference_configurations": [
|
|
50
|
-
"ON_DEMAND"
|
|
50
|
+
"ON_DEMAND",
|
|
51
|
+
"INFERENCE_PROFILE"
|
|
51
52
|
],
|
|
52
53
|
"response_streaming_supported": true,
|
|
53
54
|
"input_modalities": [
|
|
@@ -64,7 +65,7 @@
|
|
|
64
65
|
"created_at": null,
|
|
65
66
|
"display_name": "Claude 3.5 Sonnet",
|
|
66
67
|
"provider": "bedrock",
|
|
67
|
-
"context_window":
|
|
68
|
+
"context_window": 200000,
|
|
68
69
|
"max_tokens": 4096,
|
|
69
70
|
"type": "chat",
|
|
70
71
|
"family": "claude3_sonnet",
|
|
@@ -94,7 +95,7 @@
|
|
|
94
95
|
"created_at": null,
|
|
95
96
|
"display_name": "Claude 3.5 Sonnet",
|
|
96
97
|
"provider": "bedrock",
|
|
97
|
-
"context_window":
|
|
98
|
+
"context_window": 200000,
|
|
98
99
|
"max_tokens": 4096,
|
|
99
100
|
"type": "chat",
|
|
100
101
|
"family": "claude3_sonnet",
|
|
@@ -124,7 +125,7 @@
|
|
|
124
125
|
"created_at": null,
|
|
125
126
|
"display_name": "Claude 3.5 Sonnet",
|
|
126
127
|
"provider": "bedrock",
|
|
127
|
-
"context_window":
|
|
128
|
+
"context_window": 200000,
|
|
128
129
|
"max_tokens": 4096,
|
|
129
130
|
"type": "chat",
|
|
130
131
|
"family": "claude3_sonnet",
|
|
@@ -154,7 +155,7 @@
|
|
|
154
155
|
"created_at": null,
|
|
155
156
|
"display_name": "Claude 3.5 Sonnet v2",
|
|
156
157
|
"provider": "bedrock",
|
|
157
|
-
"context_window":
|
|
158
|
+
"context_window": 200000,
|
|
158
159
|
"max_tokens": 4096,
|
|
159
160
|
"type": "chat",
|
|
160
161
|
"family": "claude3_sonnet",
|
|
@@ -184,7 +185,7 @@
|
|
|
184
185
|
"created_at": null,
|
|
185
186
|
"display_name": "Claude 3.5 Sonnet v2",
|
|
186
187
|
"provider": "bedrock",
|
|
187
|
-
"context_window":
|
|
188
|
+
"context_window": 200000,
|
|
188
189
|
"max_tokens": 4096,
|
|
189
190
|
"type": "chat",
|
|
190
191
|
"family": "claude3_sonnet",
|
|
@@ -214,7 +215,7 @@
|
|
|
214
215
|
"created_at": null,
|
|
215
216
|
"display_name": "Claude 3.5 Sonnet v2",
|
|
216
217
|
"provider": "bedrock",
|
|
217
|
-
"context_window":
|
|
218
|
+
"context_window": 200000,
|
|
218
219
|
"max_tokens": 4096,
|
|
219
220
|
"type": "chat",
|
|
220
221
|
"family": "claude3_sonnet",
|
|
@@ -244,7 +245,7 @@
|
|
|
244
245
|
"created_at": null,
|
|
245
246
|
"display_name": "Claude 3.5 Sonnet v2",
|
|
246
247
|
"provider": "bedrock",
|
|
247
|
-
"context_window":
|
|
248
|
+
"context_window": 200000,
|
|
248
249
|
"max_tokens": 4096,
|
|
249
250
|
"type": "chat",
|
|
250
251
|
"family": "claude3_sonnet",
|
|
@@ -269,36 +270,6 @@
|
|
|
269
270
|
]
|
|
270
271
|
}
|
|
271
272
|
},
|
|
272
|
-
{
|
|
273
|
-
"id": "us.anthropic.claude-3-7-sonnet-20250219-v1:0",
|
|
274
|
-
"created_at": null,
|
|
275
|
-
"display_name": "Claude 3.7 Sonnet",
|
|
276
|
-
"provider": "bedrock",
|
|
277
|
-
"context_window": 4096,
|
|
278
|
-
"max_tokens": 4096,
|
|
279
|
-
"type": "chat",
|
|
280
|
-
"family": "claude3_sonnet",
|
|
281
|
-
"supports_vision": true,
|
|
282
|
-
"supports_functions": true,
|
|
283
|
-
"supports_json_mode": true,
|
|
284
|
-
"input_price_per_million": 3.0,
|
|
285
|
-
"output_price_per_million": 15.0,
|
|
286
|
-
"metadata": {
|
|
287
|
-
"provider_name": "Anthropic",
|
|
288
|
-
"customizations_supported": [],
|
|
289
|
-
"inference_configurations": [
|
|
290
|
-
"INFERENCE_PROFILE"
|
|
291
|
-
],
|
|
292
|
-
"response_streaming_supported": true,
|
|
293
|
-
"input_modalities": [
|
|
294
|
-
"TEXT",
|
|
295
|
-
"IMAGE"
|
|
296
|
-
],
|
|
297
|
-
"output_modalities": [
|
|
298
|
-
"TEXT"
|
|
299
|
-
]
|
|
300
|
-
}
|
|
301
|
-
},
|
|
302
273
|
{
|
|
303
274
|
"id": "anthropic.claude-3-haiku-20240307-v1:0",
|
|
304
275
|
"created_at": null,
|
|
@@ -607,13 +578,13 @@
|
|
|
607
578
|
"created_at": null,
|
|
608
579
|
"display_name": "Claude Instant",
|
|
609
580
|
"provider": "bedrock",
|
|
610
|
-
"context_window":
|
|
581
|
+
"context_window": 200000,
|
|
611
582
|
"max_tokens": 4096,
|
|
612
583
|
"type": "chat",
|
|
613
584
|
"family": "claude_instant",
|
|
614
|
-
"supports_vision":
|
|
615
|
-
"supports_functions":
|
|
616
|
-
"supports_json_mode":
|
|
585
|
+
"supports_vision": true,
|
|
586
|
+
"supports_functions": true,
|
|
587
|
+
"supports_json_mode": true,
|
|
617
588
|
"input_price_per_million": 0.8,
|
|
618
589
|
"output_price_per_million": 2.4,
|
|
619
590
|
"metadata": {
|
|
@@ -636,13 +607,13 @@
|
|
|
636
607
|
"created_at": null,
|
|
637
608
|
"display_name": "Claude Instant",
|
|
638
609
|
"provider": "bedrock",
|
|
639
|
-
"context_window":
|
|
610
|
+
"context_window": 200000,
|
|
640
611
|
"max_tokens": 4096,
|
|
641
612
|
"type": "chat",
|
|
642
613
|
"family": "claude_instant",
|
|
643
|
-
"supports_vision":
|
|
644
|
-
"supports_functions":
|
|
645
|
-
"supports_json_mode":
|
|
614
|
+
"supports_vision": true,
|
|
615
|
+
"supports_functions": true,
|
|
616
|
+
"supports_json_mode": true,
|
|
646
617
|
"input_price_per_million": 0.8,
|
|
647
618
|
"output_price_per_million": 2.4,
|
|
648
619
|
"metadata": {
|
|
@@ -665,13 +636,13 @@
|
|
|
665
636
|
"created_at": null,
|
|
666
637
|
"display_name": "Claude",
|
|
667
638
|
"provider": "bedrock",
|
|
668
|
-
"context_window":
|
|
639
|
+
"context_window": 200000,
|
|
669
640
|
"max_tokens": 4096,
|
|
670
641
|
"type": "chat",
|
|
671
642
|
"family": "claude2",
|
|
672
|
-
"supports_vision":
|
|
673
|
-
"supports_functions":
|
|
674
|
-
"supports_json_mode":
|
|
643
|
+
"supports_vision": true,
|
|
644
|
+
"supports_functions": true,
|
|
645
|
+
"supports_json_mode": true,
|
|
675
646
|
"input_price_per_million": 8.0,
|
|
676
647
|
"output_price_per_million": 24.0,
|
|
677
648
|
"metadata": {
|
|
@@ -694,13 +665,13 @@
|
|
|
694
665
|
"created_at": null,
|
|
695
666
|
"display_name": "Claude",
|
|
696
667
|
"provider": "bedrock",
|
|
697
|
-
"context_window":
|
|
668
|
+
"context_window": 200000,
|
|
698
669
|
"max_tokens": 4096,
|
|
699
670
|
"type": "chat",
|
|
700
671
|
"family": "claude2",
|
|
701
|
-
"supports_vision":
|
|
702
|
-
"supports_functions":
|
|
703
|
-
"supports_json_mode":
|
|
672
|
+
"supports_vision": true,
|
|
673
|
+
"supports_functions": true,
|
|
674
|
+
"supports_json_mode": true,
|
|
704
675
|
"input_price_per_million": 8.0,
|
|
705
676
|
"output_price_per_million": 24.0,
|
|
706
677
|
"metadata": {
|
|
@@ -723,13 +694,13 @@
|
|
|
723
694
|
"created_at": null,
|
|
724
695
|
"display_name": "Claude",
|
|
725
696
|
"provider": "bedrock",
|
|
726
|
-
"context_window":
|
|
697
|
+
"context_window": 200000,
|
|
727
698
|
"max_tokens": 4096,
|
|
728
699
|
"type": "chat",
|
|
729
700
|
"family": "claude2",
|
|
730
|
-
"supports_vision":
|
|
731
|
-
"supports_functions":
|
|
732
|
-
"supports_json_mode":
|
|
701
|
+
"supports_vision": true,
|
|
702
|
+
"supports_functions": true,
|
|
703
|
+
"supports_json_mode": true,
|
|
733
704
|
"input_price_per_million": 8.0,
|
|
734
705
|
"output_price_per_million": 24.0,
|
|
735
706
|
"metadata": {
|
|
@@ -752,13 +723,13 @@
|
|
|
752
723
|
"created_at": null,
|
|
753
724
|
"display_name": "Claude",
|
|
754
725
|
"provider": "bedrock",
|
|
755
|
-
"context_window":
|
|
726
|
+
"context_window": 200000,
|
|
756
727
|
"max_tokens": 4096,
|
|
757
728
|
"type": "chat",
|
|
758
729
|
"family": "claude2",
|
|
759
|
-
"supports_vision":
|
|
760
|
-
"supports_functions":
|
|
761
|
-
"supports_json_mode":
|
|
730
|
+
"supports_vision": true,
|
|
731
|
+
"supports_functions": true,
|
|
732
|
+
"supports_json_mode": true,
|
|
762
733
|
"input_price_per_million": 8.0,
|
|
763
734
|
"output_price_per_million": 24.0,
|
|
764
735
|
"metadata": {
|
|
@@ -781,13 +752,13 @@
|
|
|
781
752
|
"created_at": null,
|
|
782
753
|
"display_name": "Claude",
|
|
783
754
|
"provider": "bedrock",
|
|
784
|
-
"context_window":
|
|
755
|
+
"context_window": 200000,
|
|
785
756
|
"max_tokens": 4096,
|
|
786
757
|
"type": "chat",
|
|
787
758
|
"family": "claude2",
|
|
788
|
-
"supports_vision":
|
|
789
|
-
"supports_functions":
|
|
790
|
-
"supports_json_mode":
|
|
759
|
+
"supports_vision": true,
|
|
760
|
+
"supports_functions": true,
|
|
761
|
+
"supports_json_mode": true,
|
|
791
762
|
"input_price_per_million": 8.0,
|
|
792
763
|
"output_price_per_million": 24.0,
|
|
793
764
|
"metadata": {
|
|
@@ -810,13 +781,13 @@
|
|
|
810
781
|
"created_at": null,
|
|
811
782
|
"display_name": "Claude",
|
|
812
783
|
"provider": "bedrock",
|
|
813
|
-
"context_window":
|
|
784
|
+
"context_window": 200000,
|
|
814
785
|
"max_tokens": 4096,
|
|
815
786
|
"type": "chat",
|
|
816
787
|
"family": "claude2",
|
|
817
|
-
"supports_vision":
|
|
818
|
-
"supports_functions":
|
|
819
|
-
"supports_json_mode":
|
|
788
|
+
"supports_vision": true,
|
|
789
|
+
"supports_functions": true,
|
|
790
|
+
"supports_json_mode": true,
|
|
820
791
|
"input_price_per_million": 8.0,
|
|
821
792
|
"output_price_per_million": 24.0,
|
|
822
793
|
"metadata": {
|
|
@@ -1613,7 +1584,8 @@
|
|
|
1613
1584
|
"output_token_limit": 8192,
|
|
1614
1585
|
"supported_generation_methods": [
|
|
1615
1586
|
"generateContent",
|
|
1616
|
-
"countTokens"
|
|
1587
|
+
"countTokens",
|
|
1588
|
+
"createCachedContent"
|
|
1617
1589
|
]
|
|
1618
1590
|
}
|
|
1619
1591
|
},
|
|
@@ -1638,7 +1610,8 @@
|
|
|
1638
1610
|
"output_token_limit": 8192,
|
|
1639
1611
|
"supported_generation_methods": [
|
|
1640
1612
|
"generateContent",
|
|
1641
|
-
"countTokens"
|
|
1613
|
+
"countTokens",
|
|
1614
|
+
"createCachedContent"
|
|
1642
1615
|
]
|
|
1643
1616
|
}
|
|
1644
1617
|
},
|
|
@@ -1794,6 +1767,31 @@
|
|
|
1794
1767
|
]
|
|
1795
1768
|
}
|
|
1796
1769
|
},
|
|
1770
|
+
{
|
|
1771
|
+
"id": "gemini-2.0-flash-live-001",
|
|
1772
|
+
"created_at": null,
|
|
1773
|
+
"display_name": "Gemini 2.0 Flash 001",
|
|
1774
|
+
"provider": "gemini",
|
|
1775
|
+
"context_window": 131072,
|
|
1776
|
+
"max_tokens": 8192,
|
|
1777
|
+
"type": "chat",
|
|
1778
|
+
"family": "gemini20_flash",
|
|
1779
|
+
"supports_vision": true,
|
|
1780
|
+
"supports_functions": true,
|
|
1781
|
+
"supports_json_mode": true,
|
|
1782
|
+
"input_price_per_million": 0.1,
|
|
1783
|
+
"output_price_per_million": 0.4,
|
|
1784
|
+
"metadata": {
|
|
1785
|
+
"version": "001",
|
|
1786
|
+
"description": "Gemini 2.0 Flash 001",
|
|
1787
|
+
"input_token_limit": 131072,
|
|
1788
|
+
"output_token_limit": 8192,
|
|
1789
|
+
"supported_generation_methods": [
|
|
1790
|
+
"bidiGenerateContent",
|
|
1791
|
+
"countTokens"
|
|
1792
|
+
]
|
|
1793
|
+
}
|
|
1794
|
+
},
|
|
1797
1795
|
{
|
|
1798
1796
|
"id": "gemini-2.0-flash-thinking-exp",
|
|
1799
1797
|
"created_at": null,
|
|
@@ -1890,7 +1888,8 @@
|
|
|
1890
1888
|
"output_token_limit": 65536,
|
|
1891
1889
|
"supported_generation_methods": [
|
|
1892
1890
|
"generateContent",
|
|
1893
|
-
"countTokens"
|
|
1891
|
+
"countTokens",
|
|
1892
|
+
"createCachedContent"
|
|
1894
1893
|
]
|
|
1895
1894
|
}
|
|
1896
1895
|
},
|
|
@@ -1915,7 +1914,8 @@
|
|
|
1915
1914
|
"output_token_limit": 65536,
|
|
1916
1915
|
"supported_generation_methods": [
|
|
1917
1916
|
"generateContent",
|
|
1918
|
-
"countTokens"
|
|
1917
|
+
"countTokens",
|
|
1918
|
+
"createCachedContent"
|
|
1919
1919
|
]
|
|
1920
1920
|
}
|
|
1921
1921
|
},
|
|
@@ -1940,7 +1940,34 @@
|
|
|
1940
1940
|
"output_token_limit": 65536,
|
|
1941
1941
|
"supported_generation_methods": [
|
|
1942
1942
|
"generateContent",
|
|
1943
|
-
"countTokens"
|
|
1943
|
+
"countTokens",
|
|
1944
|
+
"createCachedContent"
|
|
1945
|
+
]
|
|
1946
|
+
}
|
|
1947
|
+
},
|
|
1948
|
+
{
|
|
1949
|
+
"id": "gemini-2.5-pro-preview-03-25",
|
|
1950
|
+
"created_at": null,
|
|
1951
|
+
"display_name": "Gemini 2.5 Pro Preview 03-25",
|
|
1952
|
+
"provider": "gemini",
|
|
1953
|
+
"context_window": 1048576,
|
|
1954
|
+
"max_tokens": 65536,
|
|
1955
|
+
"type": "chat",
|
|
1956
|
+
"family": "other",
|
|
1957
|
+
"supports_vision": true,
|
|
1958
|
+
"supports_functions": true,
|
|
1959
|
+
"supports_json_mode": true,
|
|
1960
|
+
"input_price_per_million": 0.075,
|
|
1961
|
+
"output_price_per_million": 0.3,
|
|
1962
|
+
"metadata": {
|
|
1963
|
+
"version": "2.5-preview-03-25",
|
|
1964
|
+
"description": "Gemini 2.5 Pro Preview 03-25",
|
|
1965
|
+
"input_token_limit": 1048576,
|
|
1966
|
+
"output_token_limit": 65536,
|
|
1967
|
+
"supported_generation_methods": [
|
|
1968
|
+
"generateContent",
|
|
1969
|
+
"countTokens",
|
|
1970
|
+
"createCachedContent"
|
|
1944
1971
|
]
|
|
1945
1972
|
}
|
|
1946
1973
|
},
|
|
@@ -1964,7 +1991,8 @@
|
|
|
1964
1991
|
"input_token_limit": 8192,
|
|
1965
1992
|
"output_token_limit": 1,
|
|
1966
1993
|
"supported_generation_methods": [
|
|
1967
|
-
"embedContent"
|
|
1994
|
+
"embedContent",
|
|
1995
|
+
"countTextTokens"
|
|
1968
1996
|
]
|
|
1969
1997
|
}
|
|
1970
1998
|
},
|
|
@@ -1988,7 +2016,8 @@
|
|
|
1988
2016
|
"input_token_limit": 8192,
|
|
1989
2017
|
"output_token_limit": 1,
|
|
1990
2018
|
"supported_generation_methods": [
|
|
1991
|
-
"embedContent"
|
|
2019
|
+
"embedContent",
|
|
2020
|
+
"countTextTokens"
|
|
1992
2021
|
]
|
|
1993
2022
|
}
|
|
1994
2023
|
},
|
|
@@ -2013,7 +2042,8 @@
|
|
|
2013
2042
|
"output_token_limit": 65536,
|
|
2014
2043
|
"supported_generation_methods": [
|
|
2015
2044
|
"generateContent",
|
|
2016
|
-
"countTokens"
|
|
2045
|
+
"countTokens",
|
|
2046
|
+
"createCachedContent"
|
|
2017
2047
|
]
|
|
2018
2048
|
}
|
|
2019
2049
|
},
|
|
@@ -2042,6 +2072,56 @@
|
|
|
2042
2072
|
]
|
|
2043
2073
|
}
|
|
2044
2074
|
},
|
|
2075
|
+
{
|
|
2076
|
+
"id": "gemma-3-12b-it",
|
|
2077
|
+
"created_at": null,
|
|
2078
|
+
"display_name": "Gemma 3 12B",
|
|
2079
|
+
"provider": "gemini",
|
|
2080
|
+
"context_window": 32768,
|
|
2081
|
+
"max_tokens": 8192,
|
|
2082
|
+
"type": "chat",
|
|
2083
|
+
"family": "other",
|
|
2084
|
+
"supports_vision": false,
|
|
2085
|
+
"supports_functions": false,
|
|
2086
|
+
"supports_json_mode": false,
|
|
2087
|
+
"input_price_per_million": 0.075,
|
|
2088
|
+
"output_price_per_million": 0.3,
|
|
2089
|
+
"metadata": {
|
|
2090
|
+
"version": "001",
|
|
2091
|
+
"description": null,
|
|
2092
|
+
"input_token_limit": 32768,
|
|
2093
|
+
"output_token_limit": 8192,
|
|
2094
|
+
"supported_generation_methods": [
|
|
2095
|
+
"generateContent",
|
|
2096
|
+
"countTokens"
|
|
2097
|
+
]
|
|
2098
|
+
}
|
|
2099
|
+
},
|
|
2100
|
+
{
|
|
2101
|
+
"id": "gemma-3-1b-it",
|
|
2102
|
+
"created_at": null,
|
|
2103
|
+
"display_name": "Gemma 3 1B",
|
|
2104
|
+
"provider": "gemini",
|
|
2105
|
+
"context_window": 32768,
|
|
2106
|
+
"max_tokens": 8192,
|
|
2107
|
+
"type": "chat",
|
|
2108
|
+
"family": "other",
|
|
2109
|
+
"supports_vision": false,
|
|
2110
|
+
"supports_functions": false,
|
|
2111
|
+
"supports_json_mode": false,
|
|
2112
|
+
"input_price_per_million": 0.075,
|
|
2113
|
+
"output_price_per_million": 0.3,
|
|
2114
|
+
"metadata": {
|
|
2115
|
+
"version": "001",
|
|
2116
|
+
"description": null,
|
|
2117
|
+
"input_token_limit": 32768,
|
|
2118
|
+
"output_token_limit": 8192,
|
|
2119
|
+
"supported_generation_methods": [
|
|
2120
|
+
"generateContent",
|
|
2121
|
+
"countTokens"
|
|
2122
|
+
]
|
|
2123
|
+
}
|
|
2124
|
+
},
|
|
2045
2125
|
{
|
|
2046
2126
|
"id": "gemma-3-27b-it",
|
|
2047
2127
|
"created_at": null,
|
|
@@ -2067,6 +2147,31 @@
|
|
|
2067
2147
|
]
|
|
2068
2148
|
}
|
|
2069
2149
|
},
|
|
2150
|
+
{
|
|
2151
|
+
"id": "gemma-3-4b-it",
|
|
2152
|
+
"created_at": null,
|
|
2153
|
+
"display_name": "Gemma 3 4B",
|
|
2154
|
+
"provider": "gemini",
|
|
2155
|
+
"context_window": 32768,
|
|
2156
|
+
"max_tokens": 8192,
|
|
2157
|
+
"type": "chat",
|
|
2158
|
+
"family": "other",
|
|
2159
|
+
"supports_vision": false,
|
|
2160
|
+
"supports_functions": false,
|
|
2161
|
+
"supports_json_mode": false,
|
|
2162
|
+
"input_price_per_million": 0.075,
|
|
2163
|
+
"output_price_per_million": 0.3,
|
|
2164
|
+
"metadata": {
|
|
2165
|
+
"version": "001",
|
|
2166
|
+
"description": null,
|
|
2167
|
+
"input_token_limit": 32768,
|
|
2168
|
+
"output_token_limit": 8192,
|
|
2169
|
+
"supported_generation_methods": [
|
|
2170
|
+
"generateContent",
|
|
2171
|
+
"countTokens"
|
|
2172
|
+
]
|
|
2173
|
+
}
|
|
2174
|
+
},
|
|
2070
2175
|
{
|
|
2071
2176
|
"id": "gpt-3.5-turbo",
|
|
2072
2177
|
"created_at": "2023-02-28T19:56:42+01:00",
|
|
@@ -2314,6 +2419,120 @@
|
|
|
2314
2419
|
"owned_by": "system"
|
|
2315
2420
|
}
|
|
2316
2421
|
},
|
|
2422
|
+
{
|
|
2423
|
+
"id": "gpt-4.1",
|
|
2424
|
+
"created_at": "2025-04-10T22:22:22+02:00",
|
|
2425
|
+
"display_name": "GPT-4.1",
|
|
2426
|
+
"provider": "openai",
|
|
2427
|
+
"context_window": 1047576,
|
|
2428
|
+
"max_tokens": 32768,
|
|
2429
|
+
"type": "chat",
|
|
2430
|
+
"family": "gpt41",
|
|
2431
|
+
"supports_vision": true,
|
|
2432
|
+
"supports_functions": true,
|
|
2433
|
+
"supports_json_mode": true,
|
|
2434
|
+
"input_price_per_million": 2.0,
|
|
2435
|
+
"output_price_per_million": 8.0,
|
|
2436
|
+
"metadata": {
|
|
2437
|
+
"object": "model",
|
|
2438
|
+
"owned_by": "system"
|
|
2439
|
+
}
|
|
2440
|
+
},
|
|
2441
|
+
{
|
|
2442
|
+
"id": "gpt-4.1-2025-04-14",
|
|
2443
|
+
"created_at": "2025-04-10T22:09:06+02:00",
|
|
2444
|
+
"display_name": "GPT-4.1 20250414",
|
|
2445
|
+
"provider": "openai",
|
|
2446
|
+
"context_window": 1047576,
|
|
2447
|
+
"max_tokens": 32768,
|
|
2448
|
+
"type": "chat",
|
|
2449
|
+
"family": "gpt41",
|
|
2450
|
+
"supports_vision": true,
|
|
2451
|
+
"supports_functions": true,
|
|
2452
|
+
"supports_json_mode": true,
|
|
2453
|
+
"input_price_per_million": 2.0,
|
|
2454
|
+
"output_price_per_million": 8.0,
|
|
2455
|
+
"metadata": {
|
|
2456
|
+
"object": "model",
|
|
2457
|
+
"owned_by": "system"
|
|
2458
|
+
}
|
|
2459
|
+
},
|
|
2460
|
+
{
|
|
2461
|
+
"id": "gpt-4.1-mini",
|
|
2462
|
+
"created_at": "2025-04-10T22:49:33+02:00",
|
|
2463
|
+
"display_name": "GPT-4.1 Mini",
|
|
2464
|
+
"provider": "openai",
|
|
2465
|
+
"context_window": 1047576,
|
|
2466
|
+
"max_tokens": 32768,
|
|
2467
|
+
"type": "chat",
|
|
2468
|
+
"family": "gpt41_mini",
|
|
2469
|
+
"supports_vision": true,
|
|
2470
|
+
"supports_functions": true,
|
|
2471
|
+
"supports_json_mode": true,
|
|
2472
|
+
"input_price_per_million": 0.4,
|
|
2473
|
+
"output_price_per_million": 1.6,
|
|
2474
|
+
"metadata": {
|
|
2475
|
+
"object": "model",
|
|
2476
|
+
"owned_by": "system"
|
|
2477
|
+
}
|
|
2478
|
+
},
|
|
2479
|
+
{
|
|
2480
|
+
"id": "gpt-4.1-mini-2025-04-14",
|
|
2481
|
+
"created_at": "2025-04-10T22:39:07+02:00",
|
|
2482
|
+
"display_name": "GPT-4.1 Mini 20250414",
|
|
2483
|
+
"provider": "openai",
|
|
2484
|
+
"context_window": 1047576,
|
|
2485
|
+
"max_tokens": 32768,
|
|
2486
|
+
"type": "chat",
|
|
2487
|
+
"family": "gpt41_mini",
|
|
2488
|
+
"supports_vision": true,
|
|
2489
|
+
"supports_functions": true,
|
|
2490
|
+
"supports_json_mode": true,
|
|
2491
|
+
"input_price_per_million": 0.4,
|
|
2492
|
+
"output_price_per_million": 1.6,
|
|
2493
|
+
"metadata": {
|
|
2494
|
+
"object": "model",
|
|
2495
|
+
"owned_by": "system"
|
|
2496
|
+
}
|
|
2497
|
+
},
|
|
2498
|
+
{
|
|
2499
|
+
"id": "gpt-4.1-nano",
|
|
2500
|
+
"created_at": "2025-04-10T23:48:27+02:00",
|
|
2501
|
+
"display_name": "GPT-4.1 Nano",
|
|
2502
|
+
"provider": "openai",
|
|
2503
|
+
"context_window": 1047576,
|
|
2504
|
+
"max_tokens": 32768,
|
|
2505
|
+
"type": "chat",
|
|
2506
|
+
"family": "gpt41_nano",
|
|
2507
|
+
"supports_vision": true,
|
|
2508
|
+
"supports_functions": true,
|
|
2509
|
+
"supports_json_mode": true,
|
|
2510
|
+
"input_price_per_million": 0.1,
|
|
2511
|
+
"output_price_per_million": 0.4,
|
|
2512
|
+
"metadata": {
|
|
2513
|
+
"object": "model",
|
|
2514
|
+
"owned_by": "system"
|
|
2515
|
+
}
|
|
2516
|
+
},
|
|
2517
|
+
{
|
|
2518
|
+
"id": "gpt-4.1-nano-2025-04-14",
|
|
2519
|
+
"created_at": "2025-04-10T23:37:05+02:00",
|
|
2520
|
+
"display_name": "GPT-4.1 Nano 20250414",
|
|
2521
|
+
"provider": "openai",
|
|
2522
|
+
"context_window": 1047576,
|
|
2523
|
+
"max_tokens": 32768,
|
|
2524
|
+
"type": "chat",
|
|
2525
|
+
"family": "gpt41_nano",
|
|
2526
|
+
"supports_vision": true,
|
|
2527
|
+
"supports_functions": true,
|
|
2528
|
+
"supports_json_mode": true,
|
|
2529
|
+
"input_price_per_million": 0.1,
|
|
2530
|
+
"output_price_per_million": 0.4,
|
|
2531
|
+
"metadata": {
|
|
2532
|
+
"object": "model",
|
|
2533
|
+
"owned_by": "system"
|
|
2534
|
+
}
|
|
2535
|
+
},
|
|
2317
2536
|
{
|
|
2318
2537
|
"id": "gpt-4.5-preview",
|
|
2319
2538
|
"created_at": "2025-02-27T03:24:19+01:00",
|
|
@@ -2838,6 +3057,31 @@
|
|
|
2838
3057
|
]
|
|
2839
3058
|
}
|
|
2840
3059
|
},
|
|
3060
|
+
{
|
|
3061
|
+
"id": "learnlm-2.0-flash-experimental",
|
|
3062
|
+
"created_at": null,
|
|
3063
|
+
"display_name": "LearnLM 2.0 Flash Experimental",
|
|
3064
|
+
"provider": "gemini",
|
|
3065
|
+
"context_window": 1048576,
|
|
3066
|
+
"max_tokens": 32768,
|
|
3067
|
+
"type": "chat",
|
|
3068
|
+
"family": "other",
|
|
3069
|
+
"supports_vision": true,
|
|
3070
|
+
"supports_functions": true,
|
|
3071
|
+
"supports_json_mode": true,
|
|
3072
|
+
"input_price_per_million": 0.075,
|
|
3073
|
+
"output_price_per_million": 0.3,
|
|
3074
|
+
"metadata": {
|
|
3075
|
+
"version": "2.0",
|
|
3076
|
+
"description": "LearnLM 2.0 Flash Experimental",
|
|
3077
|
+
"input_token_limit": 1048576,
|
|
3078
|
+
"output_token_limit": 32768,
|
|
3079
|
+
"supported_generation_methods": [
|
|
3080
|
+
"generateContent",
|
|
3081
|
+
"countTokens"
|
|
3082
|
+
]
|
|
3083
|
+
}
|
|
3084
|
+
},
|
|
2841
3085
|
{
|
|
2842
3086
|
"id": "o1",
|
|
2843
3087
|
"created_at": "2024-12-16T20:03:36+01:00",
|
|
@@ -3028,6 +3272,44 @@
|
|
|
3028
3272
|
"owned_by": "system"
|
|
3029
3273
|
}
|
|
3030
3274
|
},
|
|
3275
|
+
{
|
|
3276
|
+
"id": "o4-mini",
|
|
3277
|
+
"created_at": "2025-04-09T21:02:31+02:00",
|
|
3278
|
+
"display_name": "O4 Mini",
|
|
3279
|
+
"provider": "openai",
|
|
3280
|
+
"context_window": 4096,
|
|
3281
|
+
"max_tokens": 16384,
|
|
3282
|
+
"type": "chat",
|
|
3283
|
+
"family": "other",
|
|
3284
|
+
"supports_vision": false,
|
|
3285
|
+
"supports_functions": false,
|
|
3286
|
+
"supports_json_mode": false,
|
|
3287
|
+
"input_price_per_million": 0.5,
|
|
3288
|
+
"output_price_per_million": 1.5,
|
|
3289
|
+
"metadata": {
|
|
3290
|
+
"object": "model",
|
|
3291
|
+
"owned_by": "system"
|
|
3292
|
+
}
|
|
3293
|
+
},
|
|
3294
|
+
{
|
|
3295
|
+
"id": "o4-mini-2025-04-16",
|
|
3296
|
+
"created_at": "2025-04-08T19:31:46+02:00",
|
|
3297
|
+
"display_name": "O4 Mini 20250416",
|
|
3298
|
+
"provider": "openai",
|
|
3299
|
+
"context_window": 4096,
|
|
3300
|
+
"max_tokens": 16384,
|
|
3301
|
+
"type": "chat",
|
|
3302
|
+
"family": "other",
|
|
3303
|
+
"supports_vision": false,
|
|
3304
|
+
"supports_functions": false,
|
|
3305
|
+
"supports_json_mode": false,
|
|
3306
|
+
"input_price_per_million": 0.5,
|
|
3307
|
+
"output_price_per_million": 1.5,
|
|
3308
|
+
"metadata": {
|
|
3309
|
+
"object": "model",
|
|
3310
|
+
"owned_by": "system"
|
|
3311
|
+
}
|
|
3312
|
+
},
|
|
3031
3313
|
{
|
|
3032
3314
|
"id": "omni-moderation-2024-09-26",
|
|
3033
3315
|
"created_at": "2024-11-27T20:07:46+01:00",
|
|
@@ -3249,6 +3531,60 @@
|
|
|
3249
3531
|
"owned_by": "system"
|
|
3250
3532
|
}
|
|
3251
3533
|
},
|
|
3534
|
+
{
|
|
3535
|
+
"id": "us.anthropic.claude-3-7-sonnet-20250219-v1:0",
|
|
3536
|
+
"created_at": null,
|
|
3537
|
+
"display_name": "Claude 3.7 Sonnet",
|
|
3538
|
+
"provider": "bedrock",
|
|
3539
|
+
"context_window": 200000,
|
|
3540
|
+
"max_tokens": 4096,
|
|
3541
|
+
"type": "chat",
|
|
3542
|
+
"family": "claude3_sonnet",
|
|
3543
|
+
"supports_vision": true,
|
|
3544
|
+
"supports_functions": true,
|
|
3545
|
+
"supports_json_mode": true,
|
|
3546
|
+
"input_price_per_million": 3.0,
|
|
3547
|
+
"output_price_per_million": 15.0,
|
|
3548
|
+
"metadata": {
|
|
3549
|
+
"provider_name": "Anthropic",
|
|
3550
|
+
"customizations_supported": [],
|
|
3551
|
+
"inference_configurations": [
|
|
3552
|
+
"INFERENCE_PROFILE"
|
|
3553
|
+
],
|
|
3554
|
+
"response_streaming_supported": true,
|
|
3555
|
+
"input_modalities": [
|
|
3556
|
+
"TEXT",
|
|
3557
|
+
"IMAGE"
|
|
3558
|
+
],
|
|
3559
|
+
"output_modalities": [
|
|
3560
|
+
"TEXT"
|
|
3561
|
+
]
|
|
3562
|
+
}
|
|
3563
|
+
},
|
|
3564
|
+
{
|
|
3565
|
+
"id": "veo-2.0-generate-001",
|
|
3566
|
+
"created_at": null,
|
|
3567
|
+
"display_name": "Veo 2",
|
|
3568
|
+
"provider": "gemini",
|
|
3569
|
+
"context_window": 480,
|
|
3570
|
+
"max_tokens": 8192,
|
|
3571
|
+
"type": "chat",
|
|
3572
|
+
"family": "other",
|
|
3573
|
+
"supports_vision": false,
|
|
3574
|
+
"supports_functions": false,
|
|
3575
|
+
"supports_json_mode": false,
|
|
3576
|
+
"input_price_per_million": 0.075,
|
|
3577
|
+
"output_price_per_million": 0.3,
|
|
3578
|
+
"metadata": {
|
|
3579
|
+
"version": "2.0",
|
|
3580
|
+
"description": "Vertex served Veo 2 model.",
|
|
3581
|
+
"input_token_limit": 480,
|
|
3582
|
+
"output_token_limit": 8192,
|
|
3583
|
+
"supported_generation_methods": [
|
|
3584
|
+
"predictLongRunning"
|
|
3585
|
+
]
|
|
3586
|
+
}
|
|
3587
|
+
},
|
|
3252
3588
|
{
|
|
3253
3589
|
"id": "whisper-1",
|
|
3254
3590
|
"created_at": "2023-02-27T22:13:04+01:00",
|
|
@@ -42,13 +42,20 @@ module RubyLLM
|
|
|
42
42
|
|
|
43
43
|
def base_model_attributes(model_id, model, slug)
|
|
44
44
|
{
|
|
45
|
-
id: model_id,
|
|
45
|
+
id: model_id_with_prefix(model_id, model),
|
|
46
46
|
created_at: nil,
|
|
47
47
|
display_name: model['modelName'] || capabilities.format_display_name(model_id),
|
|
48
48
|
provider: slug
|
|
49
49
|
}
|
|
50
50
|
end
|
|
51
51
|
|
|
52
|
+
def model_id_with_prefix(model_id, model)
|
|
53
|
+
return model_id unless model['inferenceTypesSupported']&.include?('INFERENCE_PROFILE')
|
|
54
|
+
return model_id if model['inferenceTypesSupported']&.include?('ON_DEMAND')
|
|
55
|
+
|
|
56
|
+
"us.#{model_id}"
|
|
57
|
+
end
|
|
58
|
+
|
|
52
59
|
def capability_attributes(model_id, capabilities)
|
|
53
60
|
{
|
|
54
61
|
context_window: capabilities.context_window_for(model_id),
|
|
@@ -3,13 +3,15 @@
|
|
|
3
3
|
module RubyLLM
|
|
4
4
|
module Providers
|
|
5
5
|
module OpenAI
|
|
6
|
-
|
|
7
|
-
module Capabilities # rubocop:disable Metrics/ModuleLength
|
|
6
|
+
module Capabilities # rubocop:disable Metrics/ModuleLength,Style/Documentation
|
|
8
7
|
module_function
|
|
9
8
|
|
|
10
9
|
MODEL_PATTERNS = {
|
|
11
10
|
dall_e: /^dall-e/,
|
|
12
11
|
chatgpt4o: /^chatgpt-4o/,
|
|
12
|
+
gpt41: /^gpt-4\.1(?!-(?:mini|nano))/,
|
|
13
|
+
gpt41_mini: /^gpt-4\.1-mini/,
|
|
14
|
+
gpt41_nano: /^gpt-4\.1-nano/,
|
|
13
15
|
gpt4: /^gpt-4(?:-\d{6})?$/,
|
|
14
16
|
gpt4_turbo: /^gpt-4(?:\.5)?-(?:\d{6}-)?(preview|turbo)/,
|
|
15
17
|
gpt35_turbo: /^gpt-3\.5-turbo/,
|
|
@@ -38,8 +40,9 @@ module RubyLLM
|
|
|
38
40
|
moderation: /^(?:omni|text)-moderation/
|
|
39
41
|
}.freeze
|
|
40
42
|
|
|
41
|
-
def context_window_for(model_id) # rubocop:disable Metrics/MethodLength
|
|
43
|
+
def context_window_for(model_id) # rubocop:disable Metrics/CyclomaticComplexity,Metrics/MethodLength
|
|
42
44
|
case model_family(model_id)
|
|
45
|
+
when 'gpt41', 'gpt41_mini', 'gpt41_nano' then 1_047_576
|
|
43
46
|
when 'chatgpt4o', 'gpt4_turbo', 'gpt4o', 'gpt4o_audio', 'gpt4o_mini',
|
|
44
47
|
'gpt4o_mini_audio', 'gpt4o_mini_realtime', 'gpt4o_realtime',
|
|
45
48
|
'gpt4o_search', 'gpt4o_transcribe', 'gpt4o_mini_search', 'o1_mini' then 128_000
|
|
@@ -55,6 +58,7 @@ module RubyLLM
|
|
|
55
58
|
|
|
56
59
|
def max_tokens_for(model_id) # rubocop:disable Metrics/CyclomaticComplexity,Metrics/MethodLength
|
|
57
60
|
case model_family(model_id)
|
|
61
|
+
when 'gpt41', 'gpt41_mini', 'gpt41_nano' then 32_768
|
|
58
62
|
when 'chatgpt4o', 'gpt4o', 'gpt4o_mini', 'gpt4o_mini_search' then 16_384
|
|
59
63
|
when 'babbage', 'davinci' then 16_384 # rubocop:disable Lint/DuplicateBranch
|
|
60
64
|
when 'gpt4' then 8_192
|
|
@@ -71,15 +75,16 @@ module RubyLLM
|
|
|
71
75
|
|
|
72
76
|
def supports_vision?(model_id)
|
|
73
77
|
case model_family(model_id)
|
|
74
|
-
when 'chatgpt4o', 'gpt4', 'gpt4_turbo', 'gpt4o', 'gpt4o_mini', 'o1',
|
|
75
|
-
'moderation', 'gpt4o_search', 'gpt4o_mini_search' then true
|
|
78
|
+
when 'gpt41', 'gpt41_mini', 'gpt41_nano', 'chatgpt4o', 'gpt4', 'gpt4_turbo', 'gpt4o', 'gpt4o_mini', 'o1',
|
|
79
|
+
'o1_pro', 'moderation', 'gpt4o_search', 'gpt4o_mini_search' then true
|
|
76
80
|
else false
|
|
77
81
|
end
|
|
78
82
|
end
|
|
79
83
|
|
|
80
84
|
def supports_functions?(model_id)
|
|
81
85
|
case model_family(model_id)
|
|
82
|
-
when 'gpt4', 'gpt4_turbo', 'gpt4o', 'gpt4o_mini', 'o1', 'o1_pro',
|
|
86
|
+
when 'gpt41', 'gpt41_mini', 'gpt41_nano', 'gpt4', 'gpt4_turbo', 'gpt4o', 'gpt4o_mini', 'o1', 'o1_pro',
|
|
87
|
+
'o3_mini' then true
|
|
83
88
|
when 'chatgpt4o', 'gpt35_turbo', 'o1_mini', 'gpt4o_mini_tts',
|
|
84
89
|
'gpt4o_transcribe', 'gpt4o_search', 'gpt4o_mini_search' then false
|
|
85
90
|
else false # rubocop:disable Lint/DuplicateBranch
|
|
@@ -88,7 +93,8 @@ module RubyLLM
|
|
|
88
93
|
|
|
89
94
|
def supports_structured_output?(model_id)
|
|
90
95
|
case model_family(model_id)
|
|
91
|
-
when 'chatgpt4o', 'gpt4o', 'gpt4o_mini', 'o1', 'o1_pro',
|
|
96
|
+
when 'gpt41', 'gpt41_mini', 'gpt41_nano', 'chatgpt4o', 'gpt4o', 'gpt4o_mini', 'o1', 'o1_pro',
|
|
97
|
+
'o3_mini' then true
|
|
92
98
|
else false
|
|
93
99
|
end
|
|
94
100
|
end
|
|
@@ -98,6 +104,9 @@ module RubyLLM
|
|
|
98
104
|
end
|
|
99
105
|
|
|
100
106
|
PRICES = {
|
|
107
|
+
gpt41: { input: 2.0, output: 8.0, cached_input: 0.5 },
|
|
108
|
+
gpt41_mini: { input: 0.4, output: 1.6, cached_input: 0.1 },
|
|
109
|
+
gpt41_nano: { input: 0.1, output: 0.4 },
|
|
101
110
|
chatgpt4o: { input: 5.0, output: 15.0 },
|
|
102
111
|
gpt4: { input: 10.0, output: 30.0 },
|
|
103
112
|
gpt4_turbo: { input: 10.0, output: 30.0 },
|
|
@@ -141,6 +150,12 @@ module RubyLLM
|
|
|
141
150
|
prices[:input] || prices[:price] || default_input_price
|
|
142
151
|
end
|
|
143
152
|
|
|
153
|
+
def cached_input_price_for(model_id)
|
|
154
|
+
family = model_family(model_id).to_sym
|
|
155
|
+
prices = PRICES.fetch(family, {})
|
|
156
|
+
prices[:cached_input]
|
|
157
|
+
end
|
|
158
|
+
|
|
144
159
|
def output_price_for(model_id)
|
|
145
160
|
family = model_family(model_id).to_sym
|
|
146
161
|
prices = PRICES.fetch(family, { output: default_output_price })
|
|
@@ -200,7 +215,7 @@ module RubyLLM
|
|
|
200
215
|
end
|
|
201
216
|
|
|
202
217
|
def normalize_temperature(temperature, model_id)
|
|
203
|
-
if model_id.match?(/^o
|
|
218
|
+
if model_id.match?(/^o\d/)
|
|
204
219
|
RubyLLM.logger.debug "Model #{model_id} requires temperature=1.0, ignoring provided value"
|
|
205
220
|
1.0
|
|
206
221
|
else
|
data/lib/ruby_llm/version.rb
CHANGED
data/lib/ruby_llm.rb
CHANGED
|
@@ -30,8 +30,8 @@ module RubyLLM
|
|
|
30
30
|
class Error < StandardError; end
|
|
31
31
|
|
|
32
32
|
class << self
|
|
33
|
-
def chat(model: nil, provider: nil)
|
|
34
|
-
Chat.new(model
|
|
33
|
+
def chat(model: nil, provider: nil, assume_model_exists: false)
|
|
34
|
+
Chat.new(model:, provider:, assume_model_exists:)
|
|
35
35
|
end
|
|
36
36
|
|
|
37
37
|
def embed(...)
|
data/lib/tasks/models_docs.rake
CHANGED
|
@@ -86,10 +86,18 @@ namespace :models do # rubocop:disable Metrics/BlockLength
|
|
|
86
86
|
---
|
|
87
87
|
|
|
88
88
|
# Available Models
|
|
89
|
+
{: .no_toc }
|
|
89
90
|
|
|
90
91
|
This guide lists all models available in RubyLLM, automatically generated from the current model registry.
|
|
92
|
+
{: .fs-6 .fw-300 }
|
|
91
93
|
|
|
92
|
-
|
|
94
|
+
## Table of contents
|
|
95
|
+
{: .no_toc .text-delta }
|
|
96
|
+
|
|
97
|
+
1. TOC
|
|
98
|
+
{:toc}
|
|
99
|
+
|
|
100
|
+
---
|
|
93
101
|
|
|
94
102
|
## Contributing
|
|
95
103
|
|
|
@@ -115,6 +123,10 @@ namespace :models do # rubocop:disable Metrics/BlockLength
|
|
|
115
123
|
For more information about working with models, see the [Working with Models](/guides/models) guide.
|
|
116
124
|
|
|
117
125
|
## Models by Type
|
|
126
|
+
{: .d-inline-block }
|
|
127
|
+
|
|
128
|
+
Last updated: #{Time.now.utc.strftime('%Y-%m-%d')}
|
|
129
|
+
{: .label .label-green }
|
|
118
130
|
|
|
119
131
|
### Chat Models (#{RubyLLM.models.chat_models.count})
|
|
120
132
|
|
metadata
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: ruby_llm
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 1.
|
|
4
|
+
version: 1.2.0
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Carmine Paolino
|
|
8
8
|
autorequire:
|
|
9
9
|
bindir: bin
|
|
10
10
|
cert_chain: []
|
|
11
|
-
date: 2025-04-
|
|
11
|
+
date: 2025-04-17 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: base64
|