Gemini 2.5 Flash — Google | Modeldex

Gemini 2.5 Flash

Active

Fast and efficient multimodal model.

Updated 4 days agoStructured data from Modeldex catalog

VisionAgenticLong context

Knowledge cutoff

Jan 31, 2025(1.2 years ago)

API release

Jun 17, 2025(10 months ago)

Not enough benchmark coverage yet for an Intelligence Index — needs at least 3 results across 2 categories.

History

Gemini 2.5 Flash became available via the Google API on 2025-06-17.

Training & availability

Training data has a knowledge cutoff of 2025-01-31 — information about events after that date is unlikely to appear in the model's responses. Google has not released the underlying model weights — access is via their hosted API only.

Capabilities

Context window: 1.0M tokens.
Max output: 65K tokens.
Input modalities: text, image, audio, video.

Recommended for: vision, agentic, long-context.

Limitations

The knowledge cutoff is 14 months old — this model will not know about recent events, releases, or API changes.

Quick start

Minimal example using the google API. Copy, paste, replace the key.

from google import genai

client = genai.Client(api_key="...")
resp = client.models.generate_content(
    model="gemini-2.5-flash",
    contents="Explain quantum computing in one sentence.",
)
print(resp.text)

Providers & performance

3 providers

Multi-provider inference routes for this model — sorted by throughput. Latency is time-to-first-token; throughput is output tokens per second. Data from OpenRouter, measured over the last 30 minutes.

Provider	Throughput	Latency (TTFT)	Input $ / 1M	Output $ / 1M	Context	Quant	Supports
Google	74tok/s	685ms	$0.3	$2.5	1.0M	—	tools · json
Google	60tok/s	1.37s	$0.3	$2.5	1.0M	—	tools · json
Google AI Studio	9tok/s

Benchmarks

Benchmark	Score	Source
LongBench v2Long context	15%	Third-party llm-stats.com
SWE-bench VerifiedCoding	53.5% resolved	Third-party Papers With Code

Integrations & tooling support

Tool calling: Supported
Structured outputs: Supported

Price vs quality

Competent benchmarks

Solid benchmark performance. Pricing not publicly available — check the provider.