Gemini 3.5 Flash

by Google DeepMind

Gemini 3.5 Flash is Google's high-efficiency multimodal language model released May 19, 2026 at Google I/O 2026. It features a 1M token context window and native multimodality (text, image, video, audio), designed to outperform previous Pro-tier models in coding and agentic tasks at approximately 40% lower cost than Gemini 3.1 Pro for those use cases.

Large Language Models
Proprietary

Primary Use Cases

  • High-volume agentic workflows
  • Multimodal document analysis
  • Cost-efficient coding assistance
  • Long-context reasoning tasks
  • API-driven automation pipelines

✅ Strengths

  • Surpasses previous-generation Pro models on coding benchmarks
  • Native multimodal processing across text, image, video, audio
  • Competitive pricing vs Gemini 3.1 Pro for agentic use cases
  • Batch API at 50% discount for high-volume workloads
  • Available via Google AI Studio and Vertex AI

⚠️ Limitations

  • Significantly more expensive than previous Gemini 3 Flash ($0.50/$3.00 vs $1.50/$9.00)
  • Not cost-effective for simple summarization or classification tasks
  • Requires Google AI or Vertex AI account for access

🏆 Performance Benchmarks

Standardized benchmark scores for objective comparison

Terminal-Bench 2.1(Coding / Agentic)
76.2%

Surpasses previous-generation Pro models on this benchmark

Technical Specifications

Developer
Google DeepMind
Category
Large Language Models
Type
🧠 Foundation Model
License
UNKNOWN
Version
3.5 Flash

Performance Benchmarks

Terminal-Bench 2.176.2%

Source: View benchmark details

Trust & Privacy

Health Status
🟢 Active
Privacy Grade
Grade B
Trains on User Data
🔄 Opt-out available
Certifications
None listed

Ecosystem & Integrations

API Access
✅ Available
Chrome Extension
❌ No
Mobile App
✅ Available
Pricing Model
💰 Pay Per Use
Free Tier
✅ Available

Pricing Breakdown

Standard API

Standard pay-per-use API access

Input$1.5/M tokens
Output$9/M tokens
0
Batch API

50% discount for batch processing workloads

Input$0.75/M tokens
Output$4.5/M tokens
0