1version: 1
2refreshedAt: '2026-11-10T03:32:22.898Z'
3models:
4 anthropic/claude-fable-5.1:
5 default:
6 prompt:
7 verdict: supported
8 source: doc
9 lastChecked: '2026-09-05'
10 evidencePath: https://platform.claude.com/docs/en/api/messages
11 max_tokens:
12 verdict: supported
13 source: doc
14 lastChecked: '2026-09-05'
15 evidencePath: https://platform.claude.com/docs/en/api/messages
16 max_completion_tokens:
17 verdict: supported
18 source: prose
19 lastChecked: '2026-09-05'
20 evidencePath: ~/mnt/mdr/skills/models-litellm/assets/config.yml providerRoutes parameterMapping max_completion_tokens -> max_tokens
21 messages:
22 verdict: supported
23 source: doc
24 lastChecked: '2026-06-05'
25 evidencePath: https://docs.anthropic.com/en/api/messages
26 temperature:
27 verdict: unsupported
28 source: doc
29 lastChecked: '2026-11-10'
30 evidencePath: ~/mnt/mdr/skills/models-litellm/assets/config.yml
31 unsupportedMode: ignored
32 note: Prior doc evidence records supported (https://docs.anthropic.com/en/api/messages). LiteLLM route anthropic/claude-fable-5.1/default drops temperature via additional_drop_params; provider support does not apply to this route.
33 top_p:
34 verdict: supported
35 source: doc
36 lastChecked: '2026-06-05'
37 evidencePath: https://docs.anthropic.com/en/api/messages
38 tools:
39 verdict: supported
40 source: doc
41 lastChecked: '2026-06-05'
42 evidencePath: https://docs.anthropic.com/en/api/messages
43 thinking:
44 verdict: supported
45 source: doc
46 lastChecked: '2026-09-05'
47 evidencePath: https://platform.claude.com/docs/en/build-with-claude/adaptive-thinking
48 note: 'Claude Opus 5 and Fable 5.1 support only the adaptive object shape: thinking.type=adaptive; manual enabled/budget_tokens thinking returns HTTP 400.'
49 reasoning_effort:
50 verdict: supported
51 source: prose
52 lastChecked: '2026-10-25'
53 evidencePath: https://platform.claude.com/docs/en/build-with-claude/effort
54 type: enum
55 values:
56 - low
57 - medium
58 - high
59 - xhigh
60 - max
61 valuesComplete: true
62 note: Canonical reasoning_effort is translated by models-litellm to thinking.type=adaptive plus output_config.effort; Fable 5.1 supports all five documented effort levels.
63 google/gemini-3.0-flash:
64 default:
65 prompt:
66 verdict: supported
67 source: doc
68 lastChecked: '2026-06-05'
69 evidencePath: https://openrouter.ai/docs/api-reference/chat-completion
70 type: string
71 temperature:
72 verdict: supported
73 source: doc
74 lastChecked: '2026-06-05'
75 evidencePath: https://openrouter.ai/docs/api-reference/chat-completion
76 max_tokens:
77 verdict: supported
78 source: doc
79 lastChecked: '2026-06-05'
80 evidencePath: https://openrouter.ai/docs/api-reference/chat-completion
81 max_completion_tokens:
82 verdict: supported
83 source: prose
84 lastChecked: '2026-06-05'
85 evidencePath: ~/mnt/mdr/skills/models-catalog/assets/models.yml parameterMapping max_completion_tokens -> max_tokens
86 response_format:
87 verdict: supported
88 source: doc
89 lastChecked: '2026-06-05'
90 evidencePath: https://openrouter.ai/docs/api-reference/chat-completion
91 tools:
92 verdict: supported
93 source: doc
94 lastChecked: '2026-06-05'
95 evidencePath: https://openrouter.ai/docs/api-reference/chat-completion
96 seed:
97 verdict: supported
98 source: eval
99 lastChecked: '2026-06-05'
100 evidencePath: ~/mnt/data/models-catalog/conformance/google_gemini-3.0-flash/default/seed.json
101 type: integer
102 videoInput:
103 verdict: supported
104 source: doc
105 lastChecked: '2026-06-05'
106 evidencePath: https://openrouter.ai/docs/guides/overview/multimodal/videos
107 type: string
108 reasoning_effort:
109 verdict: supported
110 source: doc
111 lastChecked: '2026-07-13'
112 evidencePath: https://ai.google.dev/gemini-api/docs/openai#thinking
113 type: enum
114 values:
115 - minimal
116 - low
117 - medium
118 - high
119 valuesComplete: true
120 format:
121 verdict: supported
122 source: observed
123 lastChecked: '2026-08-09'
124 type: enum
125 values:
126 - jpg
127 - png
128 zai/glm-5-2:
129 default:
130 logprobs:
131 verdict: supported
132 source: eval
133 lastChecked: '2026-07-15'
134 evidencePath: ~/mnt/data/models-catalog/conformance/zai_glm-5-2/default/logprobs.json
135 type: boolean
136 top_logprobs:
137 verdict: supported
138 source: eval
139 lastChecked: '2026-07-15'
140 evidencePath: ~/mnt/data/models-catalog/conformance/zai_glm-5-2/default/top_logprobs.json
141 type: integer
142 prompt:
143 verdict: supported
144 source: prose
145 lastChecked: '2026-06-24'
146 evidencePath: ~/mnt/mdr/skills/models-catalog/assets/config.yml
147 type: string
148 note: Prompt text is supported through the LiteLLM OpenRouter chat route.
149 frequency_penalty:
150 verdict: supported
151 source: schema
152 lastChecked: '2026-07-15'
153 evidencePath: https://openrouter.ai/api/v1/models
154 include_reasoning:
155 verdict: supported
156 source: schema
157 lastChecked: '2026-07-15'
158 evidencePath: https://openrouter.ai/api/v1/models
159 logit_bias:
160 verdict: supported
161 source: schema
162 lastChecked: '2026-07-15'
163 evidencePath: https://openrouter.ai/api/v1/models
164 max_tokens:
165 verdict: supported
166 source: schema
167 lastChecked: '2026-07-15'
168 evidencePath: https://openrouter.ai/api/v1/models
169 min_p:
170 verdict: supported
171 source: schema
172 lastChecked: '2026-07-15'
173 evidencePath: https://openrouter.ai/api/v1/models
174 parallel_tool_calls:
175 verdict: supported
176 source: schema
177 lastChecked: '2026-07-15'
178 evidencePath: https://openrouter.ai/api/v1/models
179 presence_penalty:
180 verdict: supported
181 source: schema
182 lastChecked: '2026-07-15'
183 evidencePath: https://openrouter.ai/api/v1/models
184 reasoning:
185 verdict: supported
186 source: schema
187 lastChecked: '2026-07-15'
188 evidencePath: https://openrouter.ai/api/v1/models
189 reasoning_effort:
190 verdict: supported
191 source: schema
192 lastChecked: '2026-07-15'
193 evidencePath: https://openrouter.ai/api/v1/models
194 repetition_penalty:
195 verdict: supported
196 source: schema
197 lastChecked: '2026-07-15'
198 evidencePath: https://openrouter.ai/api/v1/models
199 response_format:
200 verdict: supported
201 source: schema
202 lastChecked: '2026-07-15'
203 evidencePath: https://openrouter.ai/api/v1/models
204 seed:
205 verdict: supported
206 source: schema
207 lastChecked: '2026-07-15'
208 evidencePath: https://openrouter.ai/api/v1/models
209 stop:
210 verdict: supported
211 source: schema
212 lastChecked: '2026-07-15'
213 evidencePath: https://openrouter.ai/api/v1/models
214 structured_outputs:
215 verdict: supported
216 source: schema
217 lastChecked: '2026-07-15'
218 evidencePath: https://openrouter.ai/api/v1/models
219 temperature:
220 verdict: supported
221 source: schema
222 lastChecked: '2026-07-15'
223 evidencePath: https://openrouter.ai/api/v1/models
224 tool_choice:
225 verdict: supported
226 source: schema
227 lastChecked: '2026-07-15'
228 evidencePath: https://openrouter.ai/api/v1/models
229