Realtime AICost OptimizationCoding AgentsMultimodalGeminiEvalsBenchmarksAgent runtime infrastructureMulti-Agent SystemsAgent product updatesAntigravityPricing & limits
Gemini 2.5 Flash
Best price-performance model with thinking capabilities.
A Gemini 2.5 multimodal model release optimized for low-latency, high-volume reasoning and agentic tasks.
Pricing
Model profile · Current snapshot
Input / 1M
$0.10
Output / 1M
$0.40
Blended / 1M
$0.175
Output TPS
0
TTFT (s)
0
Model Intelligence
Context window
1,048,576 tokens
Arena ranking
7
Benchmarkable
Yes
Model level
release
Intelligence Index
6.7
Math Index
35.3
MMLU Pro
0.72
GPQA
0.47
HLE
0.04
LiveCodeBench
0.4
MATH-500
0.93
AIME
0.5
AIME 2025
0.35
IFBench
0.31
LCR
0.32
TerminalBench Hard
0.02
TAU2
0.19
Recent stories
2 linked stories
releasePRIMARY2026-05-19
Gemini 3.5 Flash ships with 76.2% Terminal-Bench 2.1 and $1.50/$9 pricing
Google shipped Gemini 3.5 Flash as a GA model with 1M context, 65K max output, and stronger agentic benchmarks than Gemini 3.1 Pro. Watch task-level cost, since third-party evals show it can exceed Gemini 3.1 Pro and GPT-5.5 Medium on some jobs.
releaseSECONDARY2026-05-19
Google launches Antigravity 2.0 with CLI, SDK, and single-call Managed Agents
Google launched Antigravity 2.0 as a desktop app plus CLI/SDK stack for multi-agent workflows, and added Managed Agents to the Gemini API with persistent Linux sandboxes. Try it for agent orchestration and API-based sandboxing, but verify harness costs and runtime fit.