← Gemini 3.1 Pro · API · High thinking
Evidenceacceptedproduct_spec
Official model specification
Google AI docs — gemini-3.1-pro-preview
ev_2026_07_gem31pro_spec
Result
1048576
Confidence S
- Tested
- 2026-07-11
Configuration
How it was measured
- Exact model
- gemini-3.1-pro-preview
- Reasoning
- high
- Tool access
- tool_calling_supported
- Prompting
- n/a
Caveats
Known limitations
- Preview id may be renamed at GA freeze.
- Multimodal capability ≠ scored UI visual skill without MMMU-style evidence.
- Context window size is not long-context competence (pair with AA-LCR later).
Weighting
Evidence weight factors
The engine weights each evidence row by these factors (0–1) when fusing scores — higher is more trusted.
relevance1.00
quality0.70
recency1.00
health1.00
comparability1.00
reliability1.00
Benchmark
Official model specification
Official context window, modalities, reasoning modes, and API identifiers.
Benchmark source →