MDL
Gemini 3 Flash Preview
Verified against Artificial Analysis · Jul 13, 2026
Capability
93
percentile index
Price /M
$0.5 / $3
in / out
Context
1M
tokens
Avg score
71.5
10 benchmarks
AA Intelligence
37.8
index
Speed
202 tok/s
TTFT 5.2s
Released
Dec 17, 2025
About
Gemini 3 Flash Preview is a proprietary model in the Gemini family from Google. Benchmarked on 10 evals, averaging 71.5.
Benchmark scores · 10
avg 71.5 — higher is betterThe research behind this model
all papers mentioning it →VibeThinker-3B: Exploring the Frontier of Verifiable Reasoning in Small Language Models▲ 104 on HF · Jun 15, 2026A Matter of TASTE: Improving Coverage and Difficulty of Agent Benchmarks▲ 65 on HF · May 27, 2026S-Agent: Spatial Tool-Use Elicits Reasoning for Spatial Intelligence▲ 39 on HF · Jun 18, 2026OVO-S-Bench: A Hierarchical Benchmark for Streaming Spatial Intelligence in Multimodal LLMs▲ 31 on HF · Jun 2, 2026VisualClaw: A Real-Time, Personalized Agent for the Physical World▲ 28 on HF · Jun 15, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
Performance & price
Output speed202 tok/s
Latency (TTFT)5.2s
Price in / out$0.5 / $3 per 1M
Blended price$1.13 / 1M
Math index97.0
Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →
History
full ledger →Capability over time
Append-only ledger — every observed change to this model's numbers.
API providers
Artificial Analysis$3/M