★VALS INDEX
87.8%
+/- 3.1
Accuracy
⏱VALS INDEX
5.8s
Latency (median)
$VALS INDEX
$0.0266
Cost / Test (1P-est)
SCORE STATUSscored
TEST SETS11
RECORDED TRIALS225
CUSTOM PASS54%
Industry benchmarks
Accuracyattempts · responded
GPQA_DIAMOND
GPQA Diamond
90.0%
30 attempts · 30 responded
IFEVAL
IFEval
85.6%
30 attempts · 30 responded
Custom 9-task suite
Mean /5n=trials
CUSTOM
json_strict
★
5.00
15 trials
CUSTOM
reasoning_multistep
★
4.67
15 trials
CUSTOM
code_gen_long
★
3.65
20 trials
CUSTOM
agentic_prompt
★
3.50
20 trials
CUSTOM
summarize
★
3.35
20 trials
CUSTOM
agentic_tool_use
★
2.93
15 trials
CUSTOM
code_debug
★
2.65
20 trials
CUSTOM
refactor_existing_code
-
2.00
15 trials
CUSTOM
creative_write
-
0.40
25 trials
Cross-judge validation
m3 vs Fable-5 Δ
No judge bias detected on this model — m3 and Fable-5 agree across all tasks.
Per-task distribution
Score
All trials on this page were run through the harness. Industry suites use deterministic family-specific reference graders (no LLM judge). The custom suite is graded by consensus of minimax-m3 + claude-fable-5. See /industry/ for the full benchmark suite.