Model

Gemini 2.0 Flash Thinking (Experimental)

CLOSED

Google · Gemini 2.0 family · released Jan 21, 2025

Google's first broadly-available reasoning model (Google DeepMind's answer to OpenAI o1); free experimental Flash variant that shows its thinking process, later superseded by the Gemini 2.5/3.x thinking line.

ReasoningCodingVisionFunction callingTool useAgentic

1548.5

Elo · rank #112

Parameters

Undisclosed

Active params

Undisclosed

Context

1M tokens

Architecture

Sparse Mixture-of-Experts Transformer (params undisclosed); Flash variant fine-tuned to emit visible chain-of-thought before its final answer

License

Proprietary

Languages

—

API price (in/out)

No hosted API

Modalities

text · vision

Benchmark results

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

AIMEMath73.3%#95

best: GPT-5.2 · 100.0%

Arena EloHuman preference1384#25

best: Claude Fable 5 · 1505

GPQA DiamondReasoning74.2%#107

best: GPT-5.6 · 94.6%

MMMUVision75.4%#34

best: Claude Fable 5 · 89.3%

PutnamBenchMath0.1%#6

best: DeepSeek-Prover-V2 671B · 7.1%

Run it locally

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

Input / M tok

—

Output / M tok

—

Gemini 2.0 family

Elo progression across releases

Gemini 2.0 FlashDec 20241339.5 Gemini 2.0 Flash Thinking (Experimental)Jan 20251548.5 Gemini 2.0 Flash-LiteFeb 20251071.8 Gemini 2.0 Pro ExperimentalFeb 20251382.2

Resources

DOCShttps://ai.google.dev/gemini-api/docs/models APIAvailable via Google API

API price — · each benchmark row carries its own source badge (see methodology)