← Models & AI tech

Google

Gemini

Google's multimodal model family. Profile aggregates the documented failure-mode classes linked to its versions.

Attribution note. These are documented failure-mode classesobserved across frontier models and grounded in each finding's cited source — their attribution to this specific version is illustrative. Qlarify Labs has not independently reproduced each finding on Gemini; per-version confidence requires reproduction (VERIFICATION §2–4). Open any finding to see its source.

Report card

Auto-derived from 23 linked findings (illustrative version attributions — see note above) — worst severity per category.

Safety
Critical1×
Prompt injection
Critical1×
Hallucination
High4×
Tool use
High2×
Bias
High2×
Jailbreak
High2×
Reasoning
Medium10×
Refusal
Medium1×

Strengths

Large context handling; multimodal input; competitive reasoning.

Known weaknesses

Format-constraint violations under strict schemas, lost-in-the-middle recall degradation, and shared reasoning/arithmetic limits.

Findings (23)

Methods that surface these

Related references

Versions tracked

gemini-1.5-progemini-2.0-flash

Cite this

Qlarify Labs. (2026). Google Gemini — known weaknesses. Retrieved from https://labs.qlarify.fi/models/gemini