antigravity-cli · reasoning
ANTIGRAVITY-CLI

gemini-3.1-pro-high

Google Gemini 3.1 Pro High via official Antigravity CLI (Google AI Pro subscription)

⚖ COMPARE MODELS
VALS INDEX
71.1% +/- 31.1
Accuracy
VALS INDEX
25.5s
Latency (median)
$VALS INDEX
$2.5499
Cost / Test (1P-est)
SCORE STATUSscored
TEST SETS13
RECORDED TRIALS409
CUSTOM PASS82%

Industry benchmarks

Accuracyattempts · responded
GPQA_DIAMOND GPQA Diamond
92.0%
50 attempts · 50 responded
HUMANEVAL_PLUS HumanEval+
91.5%
164 attempts · 164 responded
IFEVAL IFEval
75.0%
50 attempts · 50 responded
MMLU_PRO MMLU-Pro
26.0%
100 attempts · 100 responded
Industry benchmark Bars = accuracy %, black cap = that score

Custom 9-task suite

Mean /5n=trials
CUSTOM code_debug
5.00
5 trials
CUSTOM code_gen_long
5.00
5 trials
CUSTOM json_strict
5.00
5 trials
CUSTOM reasoning_multistep
5.00
5 trials
CUSTOM summarize
4.60
5 trials
CUSTOM agentic_tool_use
4.40
5 trials
CUSTOM agentic_prompt
4.20
5 trials
CUSTOM refactor_existing_code -
3.40
5 trials
CUSTOM creative_write
2.00
5 trials
Custom suite · scored 0–5 by consensus (m3 + Fable-5) ★ = task scored a 5 at least once · sorted by mean score

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.