Model

Gemini 2.0 Pro Experimental

CLOSED

Google · Gemini 2.0 family · released Feb 5, 2025

Experimental-only 2.0 Pro preview with a 2M-token context window; Google's strongest 2.0-era coding/reasoning model, never GA'd.

ReasoningCodingVisionFunction callingTool useAgentic

1382.2

Elo · rank #140

Parameters

Undisclosed

Active params

Undisclosed

Context

2M tokens

Architecture

Sparse Mixture-of-Experts Transformer (params undisclosed); never reached GA

License

Proprietary

Languages

—

API price (in/out)

No hosted API

Modalities

text · vision · audio · video

Benchmark results

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

Arena EloHuman preference1380#26

best: Claude Fable 5 · 1505

BIRD-SQLCoding59.3%#2

best: Command A · 59.5%

FACTS GroundingKnowledge82.8%#6

best: Gemini 2.5 Flash-Lite (Thinking) · 86.8%

GPQA DiamondReasoning64.7%#149

best: GPT-5.6 · 94.6%

LiveCodeBenchCoding36.0%#137

best: DeepSeek-V4-Pro (Think Max) · 93.5%

MATH-500Math91.8%#41

best: GPT-5 · 99.4%

MMLU-ProKnowledge79.1%#61

best: Claude Fable 5 · 91.5%

MMMUVision72.7%#45

best: Claude Fable 5 · 89.3%

SimpleQAKnowledge44.3%#6

best: GPT-4.5 · 62.5%

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

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

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