Gemini 3.5 Flash
Open in the interactive archive →
Overview
First Gemini 3.5-family model, launched GA at Google I/O 2026. Agent-first: beats Gemini 3.1 Pro on coding/agentic benchmarks (Terminal-Bench 2.1 76.2% vs 70.3%; MCP Atlas 83.6%) at ~25-40% lower cost and ~4x output speed. 1M-token context, full multimodal input. Powers Gemini Spark personal agent.
Details
- Vendor
- Google DeepMind
- Type
- Multimodal LLM
- Release year
- 2026
- Jurisdiction
- USA / UK
- CEO
- Demis Hassabis
- Key staffers
- Demis Hassabis (CEO), Shane Legg (Co-founder), Koray Kavukcuoglu (CTO)
- Deployment
- Cloud (Gemini API + AI Studio + Vertex AI + Antigravity 2 + Gemini app)
- Open source
- No
- Free tier
- Yes
- Subscription
- API: $1.50/M input, $9.00/M output ($0.15/M cached); Google AI Pro $19.99/mo, AI Ultra $99.99/mo
- Stock status
- $400 (GOOGL, May 2026)
- X
- @GoogleDeepMind
Capabilities
AppVoice ChatPdf GenerationSpreadsheet GenerationDictationAudio Generation
AIDb Rating
AIDb v2.0 (curated): Capability 9/10 · Currency 10/10 · Access 7/10 · Significance 5/10 → 8.30
Safety notes
Use official model card/docs for model-specific risk details.
Last verified 2026-06-11 · Primary source ↗
Cite this
“Gemini 3.5 Flash.” AIDb — The AI Database. Verified 2026-06-11. https://theaidatabase.com/m/google-deepmind-gemini-3-5-flash.html
More from Google DeepMind
AlphaFold 3 · CodeGemma · Gecko · Gemini 1.0 · Gemini 1.0 Ultra · Gemini 1.5