Gemini
Google's multimodal model, and it shows. Deep integrations with Docs, Drive, and Search make it genuinely useful if your life already lives in the Google ecosystem.
Scoring · in review
Every score is hand-assigned across five axes — utility, privacy, speed, cost, transparency. This one isn't published yet because we're still using it long enough to have an honest opinion. Slow scoring is the point.
We don’t earn anything when you click this.
Who it's for
Move faster with an AI co-pilot
Offload the repetitive, automate the tedious, and focus on the creative parts of your work that actually need you.
Learn by doing, not by watching
Ask it anything, get back an explanation calibrated exactly to your level. The best tutor you've ever had doesn't charge by the hour.
The Assessment
- ✓ Multimodal
- ✓ Chat
- × Reasoning on complex multi-step problems is weaker than Claude or GPT-4o at equivalent tiers.
- × Deep Google integration is a feature for some users and a privacy concern for others — the data flow is opaque.
- × Multimodal quality is strong for images, but video understanding still lags Gemini Ultra's advertised benchmarks in practice.
Related concepts
When 'think step by step' actually earns its keep — and when it's just expensive theater.
ConceptContext WindowsWhat the model can see right now — and why the edges matter
ConceptHallucination & GroundingWhy AI models confidently make things up — and what you can actually do about it
ConceptIn-Context LearningHow models 'learn' from examples in the prompt — without changing a single weight.
ConceptMixture of ExpertsHow modern models pretend to be huge while doing the work of something smaller.
ConceptModel CardsThe documentation labs publish when they release a model — and what they leave out.
ConceptMultimodal ModelsWhen AI learned to see, listen, and read — at the same time, in the same head
ConceptRLHFHumans rate, model learns, weird things happen — the post-training that made models pleasant to talk to.
ConceptScaling LawsWhy bigger keeps working — and the question of where it stops.
ConceptStructured OutputForcing the model to fill in a shape — and why it's harder than it looks.
ConceptTransformersThe architecture that changed what AI could do with language — and then everything else
ConceptWatermarkingInvisible signatures on AI-generated text — and why most don't survive contact with reality.
- a second-opinion read of a Claude or ChatGPT answer
- long-document Q&A where context windows used to matter
What Gemini can’t do
It will answer most questions confidently. You decide which confident answers to trust.
Gemini is fluent across modalities. Fluency at this scale tempts deference, but the work of double-checking, sense-checking, and choosing which output to act on remains the human signal in the loop.
Read the essay →