Gemini 3 Flash

LLM Provider / Model · Fast frontier intelligence, near real-time coding assistance, multimodal agentic loops

At a glance

Input price	$0.15
Output price	$0.60
Price tier	Budget
Context window	1M+
Max output	64K
Context tier	500K+
Speed tier	Fast
Latency	fast
Knowledge cutoff	2025
Modality	Multimodal (vision + audio)
Model ID	gemini-3-flash
Provider	Google
HumanEval	88%
MMLU	85%
SWE-Bench	40%
Benchmark	fast tier: strong
Released	2026
Hosting	Closed/API
Capabilities	Vision, Audio, Tool use, Streaming, Structured output, Long context, Agentic coding, Near real-time

Vision, Audio, Tool use, Streaming, Structured output, Long context, Agentic coding, Near real-time

Fast frontier intelligence, near real-time coding assistance, multimodal agentic loops

Trae's built-in model menu may not include Gemini 3 Flash — Trae primarily ships with Claude/GPT/Gemini/DeepSeek.
Kiro's spec mode is tuned for Claude models — Gemini 3 Flash may work but spec-driven flows are best with Claude.
OpenAI Codex (VS Code) is tied to OpenAI's GPT-5.1 Codex family — Gemini 3 Flash is not selectable inside this integration.
Claude Code only works with Claude models — your LLM is Gemini 3 Flash.
Devin uses Cognition's proprietary model internally — your selected Gemini 3 Flash will be ignored.
OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected Gemini 3 Flash will be ignored.
The Claude Code SDK is tuned for Claude models — Gemini 3 Flash may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
Lovable uses a fixed internal model — your selected Gemini 3 Flash will be ignored.
Replit Agent 3 uses its own model selection internally — your selected Gemini 3 Flash will be ignored.

Build a full stack around Gemini 3 Flash — Flowpicker shows compatibility warnings before you commit.