Path-limited search restricting to a subdirectory
Tests ability to restrict search to a subdirectory path. Evaluates path-limited search scoping.

These commands set up the repo before the model sees the prompt. They define the starting file structure, staged changes, and Git history.

  1. 01 git init
  2. 02 git config user.email 'test@test.com'
  3. 03 git config user.name 'Test User'
  4. 04 mkdir -p src tests
  5. 05 echo 'def connect(): return True' > src/db.py
  6. 06 echo 'def test_connect(): assert connect()' > tests/test_db.py
  7. 07 echo 'def helper(): pass' > src/utils.py
  8. 08 echo 'def test_helper(): assert helper()' > tests/test_utils.py
  9. 09 git add .
  10. 10 git commit -m 'Add source and test files'
  11. 11 echo 'git grep def -- src/' > .grep_command
  12. 12 git add .grep_command
  13. 13 git commit -m 'Add grep sentinel'
Prompt
Here is the output of a git grep command limited to the src/ directory. How many function definitions appear in the output? Output ONLY the number, nothing else.
Expected
2

Scoped model quality, cost, API time, and token usage for git_grep/f008.

Loading...
Loading raw attempt evidence…
anthropic/claude-fable-5:high__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-fable-5:low__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-fable-5:max__json_schema PASS 100% 294 in → 72 out (30 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-fable-5:medium__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-fable-5:xhigh__json_schema PASS 100% 294 in → 18 out (15 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-haiku-4.5:high PASS 100% 93 in → 96 out (84 reasoning)
2
anthropic/claude-haiku-4.5:high__json_schema PASS 100% 255 in → 244 out (219 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-haiku-4.5:low PASS 100% 93 in → 115 out (104 reasoning)
2
anthropic/claude-haiku-4.5:low__json_schema PASS 100% 255 in → 123 out (112 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-haiku-4.5:medium PASS 100% 93 in → 92 out (80 reasoning)
2
anthropic/claude-haiku-4.5:medium__json_schema PASS 100% 255 in → 220 out (226 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-haiku-4.5:none PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-haiku-4.5:none__json_schema PASS 100% 225 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-haiku-4.5:xhigh PASS 100% 93 in → 99 out (86 reasoning)
2
anthropic/claude-haiku-4.5:xhigh__json_schema PASS 100% 255 in → 138 out (128 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-opus-4.6:high PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-opus-4.6:high__json_schema PASS 100% 226 in → 26 out (1 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-opus-4.6:low PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-opus-4.6:low__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.6:max PASS 100% 64 in → 83 out (70 reasoning)
2
anthropic/claude-opus-4.6:max__json_schema PASS 100% 226 in → 100 out (76 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-opus-4.6:medium PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-opus-4.6:medium__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.6:none PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-opus-4.6:none__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.6:xhigh PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-opus-4.6:xhigh__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.7:high PASS 100% 87 in → 6 out (0 reasoning)
2
anthropic/claude-opus-4.7:high__json_schema PASS 100% 299 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.7:low PASS 100% 87 in → 6 out (0 reasoning)
2
anthropic/claude-opus-4.7:low__json_schema PASS 100% 299 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.7:max PASS 100% 87 in → 6 out (0 reasoning)
2
anthropic/claude-opus-4.7:max__json_schema PASS 100% 299 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.7:medium PASS 100% 87 in → 6 out (0 reasoning)
2
anthropic/claude-opus-4.7:medium__json_schema PASS 100% 299 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.7:none PASS 100% 87 in → 6 out (0 reasoning)
2
anthropic/claude-opus-4.7:none__json_schema PASS 100% 299 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.7:xhigh PASS 100% 87 in → 6 out (0 reasoning)
2
anthropic/claude-opus-4.7:xhigh__json_schema PASS 100% 299 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.8:high PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-opus-4.8:high__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-opus-4.8:low PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-opus-4.8:low__json_schema PASS 100% 294 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.8:max PASS 100% 82 in → 174 out (51 reasoning)
2
anthropic/claude-opus-4.8:max__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-opus-4.8:medium PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-opus-4.8:medium__json_schema PASS 100% 294 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.8:none PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-opus-4.8:none__json_schema PASS 100% 294 in → 7 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-opus-4.8:xhigh PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-opus-4.8:xhigh__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-sonnet-4.6:high PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-sonnet-4.6:high__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-sonnet-4.6:low PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-sonnet-4.6:low__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-sonnet-4.6:max PASS 100% 64 in → 97 out (78 reasoning)
2
anthropic/claude-sonnet-4.6:max__json_schema PASS 100% 226 in → 99 out (77 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-sonnet-4.6:medium PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-sonnet-4.6:medium__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-sonnet-4.6:none PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-sonnet-4.6:none__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-sonnet-4.6:xhigh PASS 100% 64 in → 5 out (0 reasoning)
2
anthropic/claude-sonnet-4.6:xhigh__json_schema PASS 100% 226 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
anthropic/claude-sonnet-5:high PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-sonnet-5:high__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-sonnet-5:low PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-sonnet-5:low__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-sonnet-5:medium PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-sonnet-5:medium__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-sonnet-5:none PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-sonnet-5:none__json_schema PASS 100% 294 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
anthropic/claude-sonnet-5:xhigh PASS 100% 82 in → 3 out (0 reasoning)
2
anthropic/claude-sonnet-5:xhigh__json_schema PASS 100% 294 in → 9 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
arcee-ai/trinity-large-thinking:high PASS 100% 60 in → 196 out (192 reasoning)
2
arcee-ai/trinity-large-thinking:low PASS 100% 60 in → 74 out (70 reasoning)
2
arcee-ai/trinity-large-thinking:medium PASS 100% 60 in → 229 out (225 reasoning)
2
arcee-ai/trinity-large-thinking:xhigh PASS 100% 60 in → 210 out (206 reasoning)
2
arcee-ai/trinity-mini:high PASS 100% 60 in → 180 out (207 reasoning)
2
arcee-ai/trinity-mini:high__json_schema PASS 100% 60 in → 216 out (252 reasoning)
2
JSON Schema Structured Output
(raw) { "count":2 }
arcee-ai/trinity-mini:low PASS 100% 60 in → 310 out (358 reasoning)
2
arcee-ai/trinity-mini:low__json_schema PASS 100% 60 in → 177 out (196 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
arcee-ai/trinity-mini:medium PASS 100% 60 in → 237 out (272 reasoning)
2
arcee-ai/trinity-mini:medium__json_schema PASS 100% 60 in → 186 out (207 reasoning)
2
JSON Schema Structured Output
(raw) { "count":2 }
arcee-ai/trinity-mini:xhigh PASS 100% 60 in → 251 out (281 reasoning)
2
arcee-ai/trinity-mini:xhigh__json_schema PASS 100% 60 in → 217 out (252 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash-0731:high PASS 100% 57 in → 118 out (128 reasoning)
2
deepseek/deepseek-v4-flash-0731:high__json_schema PASS 100% 136 in → 111 out (120 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash-0731:low PASS 100% 57 in → 93 out (90 reasoning)
2
deepseek/deepseek-v4-flash-0731:low__json_schema PASS 100% 57 in → 76 out (75 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash-0731:max PASS 100% 149 in → 60 out (58 reasoning)
2
deepseek/deepseek-v4-flash-0731:max__json_schema PASS 100% 136 in → 99 out (101 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash-0731:medium PASS 100% 57 in → 79 out (77 reasoning)
2
deepseek/deepseek-v4-flash-0731:medium__json_schema PASS 100% 57 in → 68 out (64 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash-0731:none__json_schema PASS 100% 57 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash-0731:xhigh PASS 100% 57 in → 53 out (55 reasoning)
2
deepseek/deepseek-v4-flash-0731:xhigh__json_schema PASS 100% 57 in → 107 out (104 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash:high PASS 100% 57 in → 68 out (65 reasoning)
2
deepseek/deepseek-v4-flash:high__json_schema PASS 100% 57 in → 97 out (88 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash:low PASS 100% 57 in → 79 out (88 reasoning)
2
deepseek/deepseek-v4-flash:medium PASS 100% 57 in → 58 out (55 reasoning)
2
deepseek/deepseek-v4-flash:medium__json_schema PASS 100% 57 in → 77 out (78 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash:none__json_schema PASS 100% 57 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-flash:xhigh PASS 100% 136 in → 88 out (96 reasoning)
2
deepseek/deepseek-v4-flash:xhigh__json_schema PASS 100% 136 in → 94 out (94 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-pro:high PASS 100% 57 in → 135 out (152 reasoning)
2
deepseek/deepseek-v4-pro:high__json_schema PASS 100% 57 in → 10 out (7 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-pro:low PASS 100% 57 in → 97 out (94 reasoning)
2
deepseek/deepseek-v4-pro:low__json_schema PASS 100% 244 in → 217 out (209 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
deepseek/deepseek-v4-pro:medium PASS 100% 57 in → 104 out (101 reasoning)
2
deepseek/deepseek-v4-pro:medium__json_schema PASS 100% 59 in → 80 out (71 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-pro:none__json_schema PASS 100% 59 in → 11 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
deepseek/deepseek-v4-pro:xhigh PASS 100% 136 in → 111 out (108 reasoning)
2
deepseek/deepseek-v4-pro:xhigh__json_schema PASS 100% 57 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemini-3-flash-preview:high PASS 100% 56 in → 160 out (159 reasoning)
2
google/gemini-3-flash-preview:high__json_schema PASS 100% 111 in → 476 out (471 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3-flash-preview:low PASS 100% 56 in → 168 out (167 reasoning)
2
google/gemini-3-flash-preview:low__json_schema PASS 100% 111 in → 260 out (255 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3-flash-preview:medium PASS 100% 56 in → 169 out (168 reasoning)
2
google/gemini-3-flash-preview:medium__json_schema PASS 100% 111 in → 259 out (254 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3-flash-preview:xhigh PASS 100% 56 in → 166 out (165 reasoning)
2
google/gemini-3-flash-preview:xhigh__json_schema PASS 100% 111 in → 261 out (255 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-flash-lite-preview:high PASS 100% 57 in → 284 out (283 reasoning)
2
google/gemini-3.1-flash-lite-preview:high__json_schema PASS 100% 57 in → 280 out (274 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-flash-lite-preview:low PASS 100% 56 in → 141 out (140 reasoning)
2
google/gemini-3.1-flash-lite-preview:low__json_schema PASS 100% 57 in → 112 out (102 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemini-3.1-flash-lite-preview:medium PASS 100% 56 in → 153 out (152 reasoning)
2
google/gemini-3.1-flash-lite-preview:medium__json_schema PASS 100% 57 in → 251 out (245 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-flash-lite-preview:xhigh PASS 100% 57 in → 287 out (286 reasoning)
2
google/gemini-3.1-flash-lite-preview:xhigh__json_schema PASS 100% 57 in → 298 out (292 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-pro-preview:high PASS 100% 56 in → 223 out (222 reasoning)
2
google/gemini-3.1-pro-preview:high__json_schema PASS 100% 111 in → 197 out (191 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-pro-preview:low PASS 100% 56 in → 223 out (222 reasoning)
2
google/gemini-3.1-pro-preview:low__json_schema PASS 100% 111 in → 176 out (170 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-pro-preview:medium PASS 100% 56 in → 225 out (224 reasoning)
2
google/gemini-3.1-pro-preview:medium__json_schema PASS 100% 111 in → 248 out (242 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.1-pro-preview:xhigh PASS 100% 56 in → 216 out (215 reasoning)
2
google/gemini-3.1-pro-preview:xhigh__json_schema PASS 100% 111 in → 283 out (277 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash-lite:high PASS 100% 57 in → 236 out (235 reasoning)
2
google/gemini-3.5-flash-lite:high__json_schema PASS 100% 111 in → 357 out (352 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3.5-flash-lite:low PASS 100% 57 in → 94 out (93 reasoning)
2
google/gemini-3.5-flash-lite:low__json_schema PASS 100% 57 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemini-3.5-flash-lite:max PASS 100% 56 in → 243 out (242 reasoning)
2
google/gemini-3.5-flash-lite:max__json_schema PASS 100% 57 in → 318 out (312 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash-lite:medium PASS 100% 57 in → 189 out (188 reasoning)
2
google/gemini-3.5-flash-lite:medium__json_schema PASS 100% 57 in → 317 out (311 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash-lite:xhigh PASS 100% 56 in → 234 out (233 reasoning)
2
google/gemini-3.5-flash-lite:xhigh__json_schema PASS 100% 57 in → 352 out (346 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash:high PASS 100% 56 in → 216 out (215 reasoning)
2
google/gemini-3.5-flash:high__json_schema PASS 100% 111 in → 439 out (433 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash:low PASS 100% 56 in → 194 out (193 reasoning)
2
google/gemini-3.5-flash:low__json_schema PASS 100% 111 in → 131 out (125 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash:medium PASS 100% 56 in → 233 out (232 reasoning)
2
google/gemini-3.5-flash:medium__json_schema PASS 100% 111 in → 173 out (167 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.5-flash:xhigh PASS 100% 56 in → 246 out (245 reasoning)
2
google/gemini-3.5-flash:xhigh__json_schema PASS 100% 111 in → 230 out (224 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
google/gemini-3.6-flash:high PASS 100% 57 in → 205 out (204 reasoning)
2
google/gemini-3.6-flash:high__json_schema PASS 100% 57 in → 413 out (408 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3.6-flash:low PASS 100% 57 in → 90 out (89 reasoning)
2
google/gemini-3.6-flash:low__json_schema PASS 100% 57 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemini-3.6-flash:max PASS 100% 56 in → 196 out (195 reasoning)
2
google/gemini-3.6-flash:max__json_schema PASS 100% 111 in → 205 out (200 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3.6-flash:medium PASS 100% 57 in → 182 out (181 reasoning)
2
google/gemini-3.6-flash:medium__json_schema PASS 100% 111 in → 268 out (264 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemini-3.6-flash:xhigh PASS 100% 56 in → 192 out (191 reasoning)
2
google/gemini-3.6-flash:xhigh__json_schema PASS 100% 111 in → 243 out (238 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemma-4-26b-a4b-it:high PASS 100% 71 in → 161 out (136 reasoning)
2
google/gemma-4-26b-a4b-it:high__json_schema PASS 100% 71 in → 186 out (151 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-26b-a4b-it:low PASS 100% 71 in → 175 out (153 reasoning)
2
google/gemma-4-26b-a4b-it:low__json_schema PASS 100% 71 in → 192 out (159 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-26b-a4b-it:medium PASS 100% 72 in → 177 out (149 reasoning)
2
google/gemma-4-26b-a4b-it:medium__json_schema PASS 100% 72 in → 196 out (158 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-26b-a4b-it:none PASS 100% 69 in → 2 out (0 reasoning)
2
google/gemma-4-26b-a4b-it:none__json_schema PASS 100% 69 in → 11 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-26b-a4b-it:xhigh PASS 100% 72 in → 174 out (172 reasoning)
2
google/gemma-4-26b-a4b-it:xhigh__json_schema PASS 100% 71 in → 182 out (148 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-31b-it:high PASS 100% 72 in → 150 out (119 reasoning)
2
google/gemma-4-31b-it:high__json_schema PASS 100% 71 in → 131 out (107 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-31b-it:low PASS 100% 71 in → 124 out (97 reasoning)
2
google/gemma-4-31b-it:low__json_schema PASS 100% 71 in → 128 out (103 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-31b-it:medium PASS 100% 72 in → 123 out (98 reasoning)
2
google/gemma-4-31b-it:medium__json_schema PASS 100% 71 in → 163 out (128 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
google/gemma-4-31b-it:none__json_schema PASS 100% 69 in → 6 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
google/gemma-4-31b-it:xhigh PASS 100% 72 in → 117 out (92 reasoning)
2
google/gemma-4-31b-it:xhigh__json_schema PASS 100% 72 in → 170 out (132 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
ibm-granite/granite-4.1-8b PASS 100% 66 in → 2 out
2
2
JSON Schema Structured Output
(raw) { "count": 2 }
inclusionai/ling-2.6-flash PASS 100% 72 in → 2 out
2
2
JSON Schema Structured Output
(raw) { "count": 2 }
liquid/lfm-2-24b-a2b PASS 100% 65 in → 2 out
2
minimax/minimax-m2.5:high PASS 100% 90 in → 155 out (182 reasoning)
2
minimax/minimax-m2.5:high__json_schema PASS 100% 97 in → 131 out (121 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m2.5:low PASS 100% 97 in → 141 out (138 reasoning)
2
minimax/minimax-m2.5:low__json_schema PASS 100% 90 in → 124 out (135 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m2.5:medium PASS 100% 95 in → 111 out (121 reasoning)
2
minimax/minimax-m2.5:medium__json_schema PASS 100% 90 in → 102 out (109 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m2.5:xhigh PASS 100% 90 in → 131 out (146 reasoning)
2
minimax/minimax-m2.5:xhigh__json_schema PASS 100% 90 in → 109 out (116 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m2.7:high PASS 100% 92 in → 135 out (133 reasoning)
2
minimax/minimax-m2.7:high__json_schema PASS 100% 220 in → 365 out (354 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m2.7:low PASS 100% 92 in → 86 out (84 reasoning)
2
minimax/minimax-m2.7:low__json_schema PASS 100% 220 in → 263 out (255 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
minimax/minimax-m2.7:medium PASS 100% 90 in → 183 out (212 reasoning)
2
minimax/minimax-m2.7:medium__json_schema PASS 100% 220 in → 307 out (300 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
minimax/minimax-m2.7:xhigh PASS 100% 90 in → 231 out (228 reasoning)
2
minimax/minimax-m3:high PASS 100% 228 in → 78 out (84 reasoning)
2
minimax/minimax-m3:high__json_schema PASS 100% 176 in → 8 out (7 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m3:low PASS 100% 228 in → 86 out (95 reasoning)
2
minimax/minimax-m3:low__json_schema PASS 100% 176 in → 10 out (9 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
minimax/minimax-m3:medium PASS 100% 228 in → 73 out (77 reasoning)
2
minimax/minimax-m3:medium__json_schema PASS 100% 215 in → 7 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
minimax/minimax-m3:xhigh PASS 100% 228 in → 84 out (90 reasoning)
2
minimax/minimax-m3:xhigh__json_schema PASS 100% 215 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
mistralai/devstral-2512 PASS 100% 55 in → 2 out
2
mistralai/devstral-2512__json_schema PASS 100% 55 in → 7 out
2
JSON Schema Structured Output
(raw) {"count": 2}
mistralai/mistral-medium-3-5:high PASS 100% 67 in → 196 out (212 reasoning)
2
mistralai/mistral-medium-3-5:high__json_schema PASS 100% 55 in → 169 out (177 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
mistralai/mistral-medium-3-5:low PASS 100% 67 in → 222 out (251 reasoning)
2
mistralai/mistral-medium-3-5:low__json_schema PASS 100% 55 in → 157 out (172 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
mistralai/mistral-medium-3-5:medium PASS 100% 67 in → 228 out (239 reasoning)
2
mistralai/mistral-medium-3-5:medium__json_schema PASS 100% 55 in → 171 out (197 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
mistralai/mistral-medium-3-5:none PASS 100% 67 in → 2 out (0 reasoning)
2
mistralai/mistral-medium-3-5:none__json_schema PASS 100% 55 in → 7 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
mistralai/mistral-medium-3-5:xhigh PASS 100% 67 in → 245 out (284 reasoning)
2
mistralai/mistral-medium-3-5:xhigh__json_schema PASS 100% 55 in → 99 out (103 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
moonshotai/kimi-k2.6:high PASS 100% 60 in → 101 out (116 reasoning)
2
moonshotai/kimi-k2.6:high__json_schema PASS 100% 60 in → 103 out (114 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
moonshotai/kimi-k2.6:low PASS 100% 60 in → 106 out (117 reasoning)
2
moonshotai/kimi-k2.6:low__json_schema PASS 100% 59 in → 614 out (103 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
moonshotai/kimi-k2.6:medium PASS 100% 60 in → 85 out (81 reasoning)
2
moonshotai/kimi-k2.6:xhigh PASS 100% 60 in → 127 out (124 reasoning)
2
moonshotai/kimi-k2.6:xhigh__json_schema PASS 100% 60 in → 120 out (113 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
moonshotai/kimi-k2.7-code:high PASS 100% 60 in → 77 out (74 reasoning)
2
moonshotai/kimi-k2.7-code:high__json_schema PASS 100% 60 in → 23 out (15 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
moonshotai/kimi-k2.7-code:low PASS 100% 59 in → 99 out (96 reasoning)
2
moonshotai/kimi-k2.7-code:low__json_schema PASS 100% 60 in → 57 out (48 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
moonshotai/kimi-k2.7-code:medium PASS 100% 60 in → 49 out (53 reasoning)
2
moonshotai/kimi-k2.7-code:medium__json_schema PASS 100% 60 in → 92 out (84 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
moonshotai/kimi-k2.7-code:xhigh PASS 100% 59 in → 66 out (63 reasoning)
2
moonshotai/kimi-k2.7-code:xhigh__json_schema PASS 100% 141 in → 82 out (75 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
moonshotai/kimi-k3:high PASS 100% 137 in → 87 out (71 reasoning)
2
moonshotai/kimi-k3:high__json_schema PASS 100% 231 in → 163 out (143 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
moonshotai/kimi-k3:low PASS 100% 137 in → 71 out (55 reasoning)
2
moonshotai/kimi-k3:low__json_schema PASS 100% 137 in → 100 out (84 reasoning)
2
moonshotai/kimi-k3:max PASS 100% 137 in → 92 out (76 reasoning)
2
moonshotai/kimi-k3:max__json_schema PASS 100% 231 in → 66 out (45 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
moonshotai/kimi-k3:medium PASS 100% 137 in → 91 out (75 reasoning)
2
moonshotai/kimi-k3:medium__json_schema PASS 100% 137 in → 85 out (69 reasoning)
2
moonshotai/kimi-k3:xhigh PASS 100% 137 in → 86 out (70 reasoning)
2
moonshotai/kimi-k3:xhigh__json_schema PASS 100% 137 in → 68 out (52 reasoning)
2
nvidia/nemotron-3-nano-30b-a3b:high PASS 100% 68 in → 84 out (90 reasoning)
2
nvidia/nemotron-3-nano-30b-a3b:high__json_schema PASS 100% 68 in → 126 out (138 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2 }
nvidia/nemotron-3-nano-30b-a3b:low PASS 100% 68 in → 103 out (117 reasoning)
2
nvidia/nemotron-3-nano-30b-a3b:low__json_schema PASS 100% 68 in → 158 out (185 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2 }
nvidia/nemotron-3-nano-30b-a3b:medium PASS 100% 68 in → 96 out (103 reasoning)
2
nvidia/nemotron-3-nano-30b-a3b:medium__json_schema PASS 100% 68 in → 111 out (115 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
nvidia/nemotron-3-nano-30b-a3b:none PASS 100% 68 in → 2 out (0 reasoning)
2
nvidia/nemotron-3-nano-30b-a3b:none__json_schema PASS 100% 68 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
nvidia/nemotron-3-nano-30b-a3b:xhigh PASS 100% 68 in → 120 out (129 reasoning)
2
nvidia/nemotron-3-nano-30b-a3b:xhigh__json_schema PASS 100% 68 in → 108 out (116 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
nvidia/nemotron-3-super-120b-a12b:high PASS 100% 68 in → 52 out (46 reasoning)
2
nvidia/nemotron-3-super-120b-a12b:high__json_schema PASS 100% 68 in → 70 out (59 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
nvidia/nemotron-3-super-120b-a12b:low PASS 100% 68 in → 70 out (77 reasoning)
2
nvidia/nemotron-3-super-120b-a12b:low__json_schema PASS 100% 68 in → 75 out (62 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
nvidia/nemotron-3-super-120b-a12b:medium PASS 100% 68 in → 163 out (155 reasoning)
2
nvidia/nemotron-3-super-120b-a12b:medium__json_schema PASS 100% 68 in → 65 out (52 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
nvidia/nemotron-3-super-120b-a12b:none PASS 100% 68 in → 2 out (0 reasoning)
2
nvidia/nemotron-3-super-120b-a12b:none__json_schema PASS 100% 68 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
nvidia/nemotron-3-super-120b-a12b:xhigh PASS 100% 68 in → 71 out (65 reasoning)
2
nvidia/nemotron-3-super-120b-a12b:xhigh__json_schema PASS 100% 68 in → 61 out (48 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-5.4-mini:high PASS 100% 56 in → 31 out (24 reasoning)
2
openai/gpt-5.4-mini:high__json_schema PASS 100% 89 in → 158 out (143 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-mini:low PASS 100% 56 in → 27 out (20 reasoning)
2
openai/gpt-5.4-mini:low__json_schema PASS 100% 89 in → 36 out (21 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-mini:medium PASS 100% 56 in → 28 out (21 reasoning)
2
openai/gpt-5.4-mini:medium__json_schema PASS 100% 89 in → 140 out (125 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-mini:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.4-mini:none__json_schema PASS 100% 89 in → 19 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-mini:xhigh PASS 100% 56 in → 45 out (38 reasoning)
2
openai/gpt-5.4-mini:xhigh__json_schema PASS 100% 89 in → 531 out (516 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-nano:high PASS 100% 56 in → 29 out (22 reasoning)
2
openai/gpt-5.4-nano:high__json_schema PASS 100% 89 in → 39 out (24 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-nano:low PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.4-nano:low__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-nano:medium PASS 100% 56 in → 29 out (22 reasoning)
2
openai/gpt-5.4-nano:medium__json_schema PASS 100% 89 in → 34 out (19 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-nano:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.4-nano:none__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4-nano:xhigh PASS 100% 56 in → 54 out (47 reasoning)
2
openai/gpt-5.4-nano:xhigh__json_schema PASS 100% 89 in → 139 out (124 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4:high PASS 100% 56 in → 35 out (28 reasoning)
2
openai/gpt-5.4:high__json_schema PASS 100% 89 in → 248 out (233 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4:low PASS 100% 56 in → 21 out (14 reasoning)
2
openai/gpt-5.4:low__json_schema PASS 100% 89 in → 38 out (23 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4:medium PASS 100% 56 in → 36 out (29 reasoning)
2
openai/gpt-5.4:medium__json_schema PASS 100% 89 in → 38 out (23 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.4:none__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.4:xhigh PASS 100% 56 in → 109 out (102 reasoning)
2
openai/gpt-5.4:xhigh__json_schema PASS 100% 89 in → 531 out (516 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.5:high PASS 100% 56 in → 38 out (31 reasoning)
2
openai/gpt-5.5:high__json_schema PASS 100% 89 in → 126 out (111 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.5:low PASS 100% 56 in → 30 out (23 reasoning)
2
openai/gpt-5.5:low__json_schema PASS 100% 89 in → 39 out (24 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.5:medium PASS 100% 56 in → 32 out (25 reasoning)
2
openai/gpt-5.5:medium__json_schema PASS 100% 89 in → 36 out (21 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.5:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.5:none__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.5:xhigh PASS 100% 56 in → 57 out (50 reasoning)
2
openai/gpt-5.5:xhigh__json_schema PASS 100% 89 in → 131 out (116 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-luna:high PASS 100% 56 in → 28 out (21 reasoning)
2
openai/gpt-5.6-luna:high__json_schema PASS 100% 89 in → 73 out (58 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-luna:low PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-luna:low__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-luna:max PASS 100% 56 in → 28 out (21 reasoning)
2
openai/gpt-5.6-luna:max__json_schema PASS 100% 89 in → 243 out (228 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-luna:medium PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-luna:medium__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-luna:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-luna:none__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-luna:xhigh PASS 100% 56 in → 28 out (21 reasoning)
2
openai/gpt-5.6-luna:xhigh__json_schema PASS 100% 89 in → 42 out (27 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-sol:high PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-sol:high__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-sol:low PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-sol:low__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-sol:max PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-sol:max__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-sol:medium PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-sol:medium__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-sol:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-sol:none__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-sol:xhigh PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-sol:xhigh__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-terra:high PASS 100% 56 in → 17 out (10 reasoning)
2
openai/gpt-5.6-terra:high__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-terra:low PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-terra:low__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-terra:max PASS 100% 56 in → 25 out (18 reasoning)
2
openai/gpt-5.6-terra:max__json_schema PASS 100% 89 in → 60 out (45 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-terra:medium PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-terra:medium__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-terra:none PASS 100% 56 in → 5 out (0 reasoning)
2
openai/gpt-5.6-terra:none__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-5.6-terra:xhigh PASS 100% 56 in → 21 out (14 reasoning)
2
openai/gpt-5.6-terra:xhigh__json_schema PASS 100% 89 in → 13 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
openai/gpt-oss-120b:high PASS 100% 117 in → 210 out (199 reasoning)
2
openai/gpt-oss-120b:high__json_schema PASS 100% 117 in → 138 out (119 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-120b:low PASS 100% 119 in → 62 out (62 reasoning)
2
openai/gpt-oss-120b:low__json_schema PASS 100% 115 in → 35 out (19 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-120b:medium PASS 100% 104 in → 71 out (72 reasoning)
2
openai/gpt-oss-120b:medium__json_schema PASS 100% 119 in → 101 out (99 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-120b:xhigh PASS 100% 104 in → 178 out (195 reasoning)
2
openai/gpt-oss-120b:xhigh__json_schema PASS 100% 115 in → 167 out (171 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-20b:high PASS 100% 117 in → 137 out (157 reasoning)
2
openai/gpt-oss-20b:high__json_schema PASS 100% 117 in → 181 out (113 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-20b:low PASS 100% 117 in → 23 out (13 reasoning)
2
openai/gpt-oss-20b:low__json_schema PASS 100% 117 in → 39 out (19 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-20b:medium PASS 100% 117 in → 105 out (94 reasoning)
2
openai/gpt-oss-20b:medium__json_schema PASS 100% 115 in → 122 out (124 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
openai/gpt-oss-20b:xhigh PASS 100% 119 in → 319 out (371 reasoning)
2
openai/gpt-oss-20b:xhigh__json_schema PASS 100% 115 in → 309 out (348 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-m.1:high PASS 100% 64 in → 196 out (192 reasoning)
2
poolside/laguna-m.1:high__json_schema PASS 100% 64 in → 144 out (133 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-m.1:low PASS 100% 64 in → 232 out (228 reasoning)
2
poolside/laguna-m.1:low__json_schema PASS 100% 64 in → 161 out (153 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-m.1:medium PASS 100% 64 in → 141 out (137 reasoning)
2
poolside/laguna-m.1:medium__json_schema PASS 100% 64 in → 314 out (303 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-m.1:none PASS 100% 64 in → 3 out (0 reasoning)
2
poolside/laguna-m.1:none__json_schema PASS 100% 64 in → 7 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-m.1:xhigh PASS 100% 64 in → 225 out (221 reasoning)
2
poolside/laguna-m.1:xhigh__json_schema PASS 100% 64 in → 237 out (226 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-xs-2.1:high PASS 100% 64 in → 243 out (241 reasoning)
2
poolside/laguna-xs-2.1:high__json_schema PASS 100% 64 in → 245 out (233 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-xs-2.1:low PASS 100% 64 in → 306 out (304 reasoning)
2
poolside/laguna-xs-2.1:low__json_schema PASS 100% 64 in → 269 out (258 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-xs-2.1:medium PASS 100% 64 in → 226 out (224 reasoning)
2
poolside/laguna-xs-2.1:medium__json_schema PASS 100% 64 in → 221 out (209 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-xs-2.1:none__json_schema PASS 100% 64 in → 7 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-xs-2.1:xhigh PASS 100% 64 in → 134 out (132 reasoning)
2
poolside/laguna-xs-2.1:xhigh__json_schema PASS 100% 64 in → 167 out (155 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-xs.2:high PASS 100% 101 in → 111 out (107 reasoning)
2
poolside/laguna-xs.2:high__json_schema PASS 100% 101 in → 119 out (107 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
poolside/laguna-xs.2:low PASS 100% 101 in → 124 out (120 reasoning)
2
poolside/laguna-xs.2:low__json_schema PASS 100% 101 in → 114 out (106 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-xs.2:medium PASS 100% 101 in → 115 out (111 reasoning)
2
poolside/laguna-xs.2:medium__json_schema PASS 100% 101 in → 236 out (228 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-xs.2:none PASS 100% 101 in → 3 out (0 reasoning)
2
poolside/laguna-xs.2:none__json_schema PASS 100% 101 in → 7 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
poolside/laguna-xs.2:xhigh PASS 100% 101 in → 122 out (118 reasoning)
2
poolside/laguna-xs.2:xhigh__json_schema PASS 100% 101 in → 205 out (197 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
qwen/qwen3.6-27b:high PASS 100% 60 in → 295 out (294 reasoning)
2
qwen/qwen3.6-27b:low PASS 100% 60 in → 245 out (239 reasoning)
2
qwen/qwen3.6-27b:low__json_schema PASS 100% 60 in → 541 out (531 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-27b:medium PASS 100% 60 in → 402 out (397 reasoning)
2
qwen/qwen3.6-27b:none PASS 100% 62 in → 2 out (0 reasoning)
2
qwen/qwen3.6-27b:none__json_schema PASS 100% 62 in → 11 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-27b:xhigh PASS 100% 60 in → 338 out (311 reasoning)
2
qwen/qwen3.6-27b:xhigh__json_schema PASS 100% 60 in → 454 out (441 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-35b-a3b:high PASS 100% 60 in → 592 out (586 reasoning)
2
qwen/qwen3.6-35b-a3b:high__json_schema PASS 100% 60 in → 145 out (146 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-35b-a3b:low PASS 100% 60 in → 446 out (419 reasoning)
2
qwen/qwen3.6-35b-a3b:low__json_schema PASS 100% 60 in → 186 out (177 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-35b-a3b:medium PASS 100% 60 in → 239 out (234 reasoning)
2
qwen/qwen3.6-35b-a3b:medium__json_schema PASS 100% 60 in → 201 out (189 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-35b-a3b:none PASS 100% 62 in → 2 out (0 reasoning)
2
qwen/qwen3.6-35b-a3b:none__json_schema PASS 100% 62 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-35b-a3b:xhigh PASS 100% 60 in → 162 out (168 reasoning)
2
qwen/qwen3.6-35b-a3b:xhigh__json_schema PASS 100% 60 in → 188 out (178 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
qwen/qwen3.6-flash:high PASS 100% 60 in → 329 out (324 reasoning)
2
qwen/qwen3.6-flash:low PASS 100% 60 in → 253 out (247 reasoning)
2
qwen/qwen3.6-flash:medium PASS 100% 60 in → 149 out (143 reasoning)
2
qwen/qwen3.6-flash:none PASS 100% 62 in → 1 out (0 reasoning)
2
qwen/qwen3.6-flash:xhigh PASS 100% 60 in → 171 out (166 reasoning)
2
qwen/qwen3.7-flash:high PASS 100% 60 in → 152 out (147 reasoning)
2
qwen/qwen3.7-flash:low PASS 100% 60 in → 115 out (110 reasoning)
2
qwen/qwen3.7-flash:max PASS 100% 60 in → 107 out (101 reasoning)
2
qwen/qwen3.7-flash:medium PASS 100% 60 in → 173 out (168 reasoning)
2
qwen/qwen3.7-flash:none PASS 100% 62 in → 1 out (0 reasoning)
2
qwen/qwen3.7-flash:none__json_schema PASS 100% 62 in → 1 out (0 reasoning)
2
qwen/qwen3.7-flash:xhigh PASS 100% 60 in → 259 out (253 reasoning)
2
qwen/qwen3.7-max:high PASS 100% 60 in → 352 out (346 reasoning)
2
qwen/qwen3.7-max:low PASS 100% 60 in → 321 out (315 reasoning)
2
qwen/qwen3.7-max:medium PASS 100% 60 in → 309 out (303 reasoning)
2
qwen/qwen3.7-max:none PASS 100% 62 in → 1 out (0 reasoning)
2
qwen/qwen3.7-max:xhigh PASS 100% 60 in → 264 out (259 reasoning)
2
qwen/qwen3.7-plus:high PASS 100% 60 in → 287 out (282 reasoning)
2
qwen/qwen3.7-plus:low PASS 100% 60 in → 207 out (201 reasoning)
2
qwen/qwen3.7-plus:medium PASS 100% 60 in → 178 out (172 reasoning)
2
qwen/qwen3.7-plus:none PASS 100% 62 in → 1 out (0 reasoning)
2
qwen/qwen3.7-plus:xhigh PASS 100% 60 in → 284 out (279 reasoning)
2
tencent/hy3:high PASS 100% 64 in → 152 out (149 reasoning)
2
tencent/hy3:high__json_schema PASS 100% 64 in → 155 out (155 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
tencent/hy3:low PASS 100% 64 in → 98 out (95 reasoning)
2
tencent/hy3:low__json_schema PASS 100% 64 in → 139 out (148 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
tencent/hy3:medium PASS 100% 64 in → 149 out (160 reasoning)
2
tencent/hy3:medium__json_schema PASS 100% 64 in → 112 out (109 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
tencent/hy3:none PASS 100% 67 in → 2 out (0 reasoning)
2
tencent/hy3:none__json_schema PASS 100% 67 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
tencent/hy3:xhigh PASS 100% 64 in → 164 out (161 reasoning)
2
tencent/hy3:xhigh__json_schema PASS 100% 64 in → 100 out (103 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
thinkingmachines/inkling-small:high PASS 100% 65 in → 44 out (37 reasoning)
2
thinkingmachines/inkling-small:high__json_schema PASS 100% 65 in → 67 out (56 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
thinkingmachines/inkling-small:low PASS 100% 65 in → 30 out (24 reasoning)
2
thinkingmachines/inkling-small:low__json_schema PASS 100% 65 in → 39 out (25 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling-small:max PASS 100% 65 in → 24 out (17 reasoning)
2
thinkingmachines/inkling-small:max__json_schema PASS 100% 65 in → 55 out (41 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling-small:medium PASS 100% 65 in → 43 out (39 reasoning)
2
thinkingmachines/inkling-small:medium__json_schema PASS 100% 65 in → 29 out (17 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling-small:xhigh PASS 100% 65 in → 45 out (40 reasoning)
2
thinkingmachines/inkling-small:xhigh__json_schema PASS 100% 65 in → 49 out (35 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling:high PASS 100% 65 in → 57 out (56 reasoning)
2
thinkingmachines/inkling:low PASS 100% 65 in → 28 out (21 reasoning)
2
thinkingmachines/inkling:low__json_schema PASS 100% 65 in → 34 out (20 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling:max PASS 100% 65 in → 111 out (103 reasoning)
2
thinkingmachines/inkling:max__json_schema PASS 100% 65 in → 88 out (76 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling:medium PASS 100% 65 in → 96 out (88 reasoning)
2
thinkingmachines/inkling:medium__json_schema PASS 100% 65 in → 67 out (53 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling:none__json_schema PASS 100% 63 in → 10 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
thinkingmachines/inkling:xhigh PASS 100% 65 in → 97 out (95 reasoning)
2
thinkingmachines/inkling:xhigh__json_schema PASS 100% 65 in → 107 out (93 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
x-ai/grok-4.3:high PASS 100% 237 in → 163 out (162 reasoning)
2
x-ai/grok-4.3:high__json_schema PASS 100% 291 in → 266 out (261 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
x-ai/grok-4.3:low__json_schema PASS 100% 297 in → 254 out (249 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
x-ai/grok-4.3:max PASS 100% 237 in → 183 out (182 reasoning)
2
x-ai/grok-4.3:max__json_schema PASS 100% 291 in → 269 out (264 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
x-ai/grok-4.3:medium__json_schema PASS 100% 297 in → 284 out (279 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
x-ai/grok-4.3:none PASS 100% 235 in → 1 out (0 reasoning)
2
x-ai/grok-4.3:none__json_schema PASS 100% 289 in → 5 out (0 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
x-ai/grok-4.3:xhigh PASS 100% 237 in → 176 out (175 reasoning)
2
x-ai/grok-4.3:xhigh__json_schema PASS 100% 291 in → 260 out (255 reasoning)
2
JSON Schema Structured Output
(raw) {"count":2}
x-ai/grok-4.5:high PASS 100% 260 in → 154 out (153 reasoning)
2
x-ai/grok-4.5:high__json_schema PASS 100% 324 in → 216 out (210 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
x-ai/grok-4.5:low PASS 100% 260 in → 114 out (113 reasoning)
2
x-ai/grok-4.5:low__json_schema PASS 100% 324 in → 281 out (275 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
x-ai/grok-4.5:max PASS 100% 260 in → 147 out (146 reasoning)
2
x-ai/grok-4.5:max__json_schema PASS 100% 324 in → 279 out (273 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
x-ai/grok-4.5:medium PASS 100% 260 in → 177 out (176 reasoning)
2
x-ai/grok-4.5:medium__json_schema PASS 100% 324 in → 310 out (304 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
x-ai/grok-4.5:xhigh PASS 100% 260 in → 135 out (134 reasoning)
2
x-ai/grok-4.5:xhigh__json_schema PASS 100% 324 in → 328 out (322 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
z-ai/glm-4.7-flash:high PASS 100% 55 in → 272 out (271 reasoning)
2
z-ai/glm-4.7-flash:high__json_schema PASS 100% 55 in → 279 out (265 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
z-ai/glm-4.7-flash:low PASS 100% 55 in → 273 out (265 reasoning)
2
z-ai/glm-4.7-flash:medium PASS 100% 55 in → 240 out (220 reasoning)
2
z-ai/glm-4.7-flash:none__json_schema PASS 100% 55 in → 8 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
z-ai/glm-4.7-flash:xhigh PASS 100% 55 in → 236 out (233 reasoning)
2
z-ai/glm-5.2:high PASS 100% 62 in → 73 out (71 reasoning)
2
z-ai/glm-5.2:high__json_schema PASS 100% 235 in → 192 out (184 reasoning)
2
JSON Schema Structured Output
(raw) {"count": 2}
z-ai/glm-5.2:low PASS 100% 62 in → 231 out (225 reasoning)
2
z-ai/glm-5.2:low__json_schema PASS 100% 62 in → 222 out (209 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
z-ai/glm-5.2:medium PASS 100% 62 in → 219 out (0 reasoning)
2
z-ai/glm-5.2:none PASS 100% 56 in → 2 out (0 reasoning)
2
z-ai/glm-5.2:none__json_schema PASS 100% 56 in → 12 out (0 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
z-ai/glm-5.2:xhigh PASS 100% 62 in → 212 out (209 reasoning)
2
z-ai/glm-5.2:xhigh__json_schema PASS 100% 62 in → 223 out (210 reasoning)
2
JSON Schema Structured Output
(raw) { "count": 2 }
anthropic/claude-fable-5:high FAIL 0% 82 in → 4 out (0 reasoning)
(empty output)
Failure: Expected '2', got ''
anthropic/claude-fable-5:low FAIL 0% 82 in → 2 out (0 reasoning)
(empty output)
Failure: Expected '2', got ''
anthropic/claude-fable-5:max FAIL 0% 82 in → 4 out (0 reasoning)
(empty output)
Failure: Expected '2', got ''
anthropic/claude-fable-5:medium FAIL 0% 82 in → 1 out (0 reasoning)
(empty output)
Failure: Expected '2', got ''
anthropic/claude-fable-5:xhigh FAIL 0% 82 in → 4 out (0 reasoning)
(empty output)
Failure: Expected '2', got ''
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
deepseek/deepseek-v4-flash-0731:none FAIL 0% 57 in → 2 out (0 reasoning)
5
Failure: Expected '2', got '5'
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
deepseek/deepseek-v4-flash:none FAIL 0% 57 in → 2 out (0 reasoning)
6
Failure: Expected '2', got '6'
deepseek/deepseek-v4-pro:none FAIL 0% 57 in → 2 out (0 reasoning)
4
Failure: Expected '2', got '4'
(empty output)
Failure: Model 'google/gemma-4-31b-it:none' violated the reasoning_level=none invariant: provider reported 109 reasoning token(s)
Invalid structured output. Output: 2
JSON Schema Structured Output
Structured Output Error
Structured output schema validation failed: $ must be of type object
Failure: Structured output schema validation failed: $ must be of type object
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
poolside/laguna-xs-2.1:none FAIL 0% 64 in → 49 out (0 reasoning)
2 The output has 2 function definitions. Each line in the git grep output shows a function definition from different files (src/db.py and src/utils.py), so there are 2 total. The answer is 2.</think>2
Failure: Expected '2', got '2 The output has 2 function definitions. Each line in the git grep output shows a function definition from different files (src/db.py and src/utils.py), so there are 2 total. The answer is 2.</think>2'
Invalid structured output. Output: 2
JSON Schema Structured Output
Structured Output Error
Structured output schema validation failed: $ must be of type object
Failure: Structured output schema validation failed: $ must be of type object
Invalid structured output. Output: 2
JSON Schema Structured Output
Structured Output Error
Structured output schema validation failed: $ must be of type object
Failure: Structured output schema validation failed: $ must be of type object
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-a3da1860-6f7b-9fa3-96b0-a3f8647b76ba","request_id":"a3da1860-6f7b-9fa3-96b0-a3f8647b76ba"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-468d8da4-b6cd-9dd9-a2dd-5f6ef56d0046","request_id":"468d8da4-b6cd-9dd9-a2dd-5f6ef56d0046"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-28c73d98-2fcc-9119-a0a8-566b9f963a05","request_id":"28c73d98-2fcc-9119-a0a8-566b9f963a05"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-1b0e01b5-b135-9e20-9fa1-910aae9c16bb","request_id":"1b0e01b5-b135-9e20-9fa1-910aae9c16bb"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-d2a2c3ae-8c49-9f3a-bc1a-9f6f73f1e0be","request_id":"d2a2c3ae-8c49-9f3a-bc1a-9f6f73f1e0be"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': 'data: {"error":{"code":"invalid_parameter_error","param":null,"message":"\'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error"},"id":"chatcmpl-a5d7a4a4-898d-98fb-976a-fec3e2c985bb"}\n\n', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': 'data: {"error":{"code":"invalid_parameter_error","param":null,"message":"\'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error"},"id":"chatcmpl-dbb83aba-037f-9d5e-85b7-c6f7e093f81d"}\n\n', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': 'data: {"error":{"code":"invalid_parameter_error","param":null,"message":"\'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error"},"id":"chatcmpl-ab5f3678-d88f-90ef-91a8-6b6da2662cf3"}\n\n', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': 'data: {"error":{"code":"invalid_parameter_error","param":null,"message":"\'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error"},"id":"chatcmpl-e0d9ac35-9736-9f04-94df-357c0238fdf2"}\n\n', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': 'data: {"error":{"code":"invalid_parameter_error","param":null,"message":"\'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error"},"id":"chatcmpl-51daf42c-ede1-95db-b83e-469bc3845763"}\n\n', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-e1cf98d3-0c19-998b-9b94-27120002b718","request_id":"e1cf98d3-0c19-998b-9b94-27120002b718"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-9811e528-a176-96f3-987a-68b99b8e9032","request_id":"9811e528-a176-96f3-987a-68b99b8e9032"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-d56622b5-5eee-9528-844c-7a0029c33de6","request_id":"d56622b5-5eee-9528-844c-7a0029c33de6"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-627beb9e-8afe-9442-9b07-4912f5349753","request_id":"627beb9e-8afe-9442-9b07-4912f5349753"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-10d935be-aade-9824-99fd-412000996950","request_id":"10d935be-aade-9824-99fd-412000996950"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-24e96fec-3808-9f66-ab75-19a43277b34e","request_id":"24e96fec-3808-9f66-ab75-19a43277b34e"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-7ac1b8cb-a2a6-9c02-a649-19c863c8a036","request_id":"7ac1b8cb-a2a6-9c02-a649-19c863c8a036"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-dc47ed38-0d3b-9e3e-9713-3e8763191079","request_id":"dc47ed38-0d3b-9e3e-9713-3e8763191079"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-389f0814-97ca-948d-a855-0b98bbbaf8e5","request_id":"389f0814-97ca-948d-a855-0b98bbbaf8e5"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
(empty output)
JSON Schema Structured Output
Failure: Error code: 400 - {'error': {'message': 'Provider returned error', 'code': 400, 'metadata': {'raw': '{"error":{"message":"<400> InternalError.Algo.InvalidParameter: \'messages\' must contain the word \'json\' in some form, to use \'response_format\' of type \'json_object\'.","type":"invalid_request_error","param":null,"code":"invalid_parameter_error"},"id":"chatcmpl-a4b6aae1-d0c7-9eac-bbb5-5961734f1863","request_id":"a4b6aae1-d0c7-9eac-bbb5-5961734f1863"}', 'provider_name': 'Alibaba', 'is_byok': False}}, 'user_id': 'user_3FrVBkmFqVIuZ58dEh7AU8vhXUV'}
thinkingmachines/inkling-small:none FAIL 0% 65 in → 4 out (0 reasoning)
2<|end_message|>
Failure: Expected '2', got '2<|end_message|>'
Invalid JSON. Output: 2{ "count": 2 }
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Extra data: line 1 column 2 (char 1)
Failure: Failed to parse structured JSON response: Extra data: line 1 column 2 (char 1)
Invalid structured output. Output: 2
JSON Schema Structured Output
Structured Output Error
Structured output schema validation failed: $ must be of type object
Failure: Structured output schema validation failed: $ must be of type object
thinkingmachines/inkling:none FAIL 0% 65 in → 4 out (0 reasoning)
2<|end_message|>
Failure: Expected '2', got '2<|end_message|>'
x-ai/grok-4.3:low FAIL 0% 243 in → 246 out (183 reasoning)
2 The provided git grep output contains exactly two lines, and each line matches a single Python function definition (one for `connect` in db.py, and one for `helper` in utils.py). No additional `def` statements or function signatures are present. Therefore the count of function definitions is 2.
Failure: Expected '2', got '2 The provided git grep output contains exactly two lines, and each line matches a single Python function definition (one for `connect` in db.py, and one for `helper` in utils.py). No additional `def` statements or function signatures are present. Therefore the count of function definitions is 2.'
x-ai/grok-4.3:medium FAIL 0% 243 in → 245 out (217 reasoning)
2 (The provided git grep output contains exactly two matches for function definitions, each introduced by the keyword "def" on its own line.)
Failure: Expected '2', got '2 (The provided git grep output contains exactly two matches for function definitions, each introduced by the keyword "def" on its own line.)'
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
z-ai/glm-4.7-flash:none FAIL 0% 55 in → 2 out (0 reasoning)
3
Failure: Expected '2', got '3'
Invalid JSON. Output:
JSON Schema Structured Output
Structured Output Error
Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
Failure: Failed to parse structured JSON response: Expecting value: line 1 column 1 (char 0)
z-ai/glm-5.2:medium__json_schema FAIL 0% 62 in → 212 out (206 reasoning)
0
JSON Schema Structured Output
(raw) { "count": 0 }
Failure: Expected '2', got '0'