Model

Gemini 1.5 Flash-8B

CLOSED

Google · Gemini 1.5 family · released Oct 3, 2024

Smallest, cheapest 1.5 Flash tier for high-volume tasks; ~80-90% of Flash's quality at a fraction of the cost.

ReasoningCodingVisionFunction callingTool useAgentic

532.6

Elo · rank #294

Parameters

Active params

8B (dense)

Context

1M tokens

Architecture

Dense Transformer decoder model (single-digit-billion params per Google), inherits Gemini 1.5 Flash's architecture and optimizations

License

Proprietary

Languages

—

API price (in/out)

$0.0375 / $0.15

Modalities

text · vision · audio · video

Benchmark results

Bar shows position within the tracked field; marker = field best

BIG-Bench HardReasoning69.5%#59

best: ERNIE 4.5 300B-A47B · 94.3%

DocVQAVision73.6%#44

best: Qwen2-VL-72B · 96.5%

GPQA DiamondReasoning30.8%#256

best: GPT-5.6 · 94.6%

MATH-500Math35.9%#151

best: GPT-5 · 99.4%

MGSMMath70.5%#34

best: OpenAI o4-mini · 93.7%

MMLUKnowledge68.1%#159

best: OpenAI o3 · 92.9%

MMMUVision50.3%#95

best: Claude Fable 5 · 89.3%

TextVQAVision66.7%#24

best: Molmo 2 8B · 85.7%

Run it locally

Closed weights — available via API only. No local deployment.

Input / M tok

$0.0375

Output / M tok

$0.15

Gemini 1.5 family

Elo progression across releases

Gemini 1.5 ProFeb 20241062.3 Gemini 1.5 FlashMay 2024865.5 Gemini 1.5 Flash-8BOct 2024532.6

Resources

DOCSai.google.dev/gemini-api/docs/models APIAvailable via Google API

API price $0.0375/$0.15 · each benchmark row carries its own source badge (see methodology)