Gemini 3 Flash
Frontier intelligence built for speed at a fraction of the cost. Achieves 90.4% on GPQA Diamond and 81.2% on MMMU Pro with Pro-grade reasoning at Flash-level latency. 3x faster than 2.5 Pro while using 30% fewer tokens on average. Ideal for agentic workflows, coding (78% on SWE-bench Verified), and multimodal reasoning with 1M token context. If you want to compare the best LLMs for your data, try Agentset.
Model Information
- Provider
- License
- Proprietary
- Input Price per 1M
- $0.50
- Output Price per 1M
- $3.00
- Context Window
- 1049K
- Release Date
- 2025-12-17
- Model Name
- gemini-3-flash-preview
- Total Evaluations
- 1350
Performance Record
LLMs Are Just One Piece of RAG
Agentset gives you a managed RAG pipeline with the top-ranked models and best practices baked in. No infrastructure to maintain, no LLM orchestration to manage.
Trusted by teams building production RAG applications
Performance Overview
ELO ratings by dataset
Gemini 3 Flash's ELO performance varies across different benchmark datasets, showing its strengths in specific domains.
Gemini 3 Flash - ELO by Dataset
Detailed Metrics
Dataset breakdown
Performance metrics across different benchmark datasets, including accuracy and latency percentiles.
PG
Quality Metrics
- Correctness
- 5.00
- Faithfulness
- 5.00
- Grounding
- 5.00
- Relevance
- 5.00
- Completeness
- 5.00
- Overall
- 5.00
Latency Distribution
- Mean
- 9444ms
- Min
- 5346ms
- Max
- 12549ms
MSMARCO
Quality Metrics
- Correctness
- 4.80
- Faithfulness
- 4.83
- Grounding
- 4.83
- Relevance
- 5.00
- Completeness
- 4.87
- Overall
- 4.87
Latency Distribution
- Mean
- 6852ms
- Min
- 3389ms
- Max
- 9837ms
SciFact
Quality Metrics
- Correctness
- 5.00
- Faithfulness
- 5.00
- Grounding
- 5.00
- Relevance
- 4.97
- Completeness
- 4.87
- Overall
- 4.97
Latency Distribution
- Mean
- 7110ms
- Min
- 3784ms
- Max
- 18224ms
Build RAG in Minutes, Not Months
Agentset gives you a complete RAG API with top-ranked LLMs and smart retrieval built in. Upload your data, call the API, and get grounded answers from day one.
import { Agentset } from "agentset";
const agentset = new Agentset();
const ns = agentset.namespace("ns_1234");
const results = await ns.search(
"What is multi-head attention?"
);
for (const result of results) {
console.log(result.text);
}Compare Models
See how it stacks up
Compare Gemini 3 Flash with other top llms to understand the differences in performance, accuracy, and latency.