Gemini 2.5 Flash

2.5 Flash

Fast multimodal Google model from the Gemini 2.5 family with hybrid reasoning and a 1M-token context window, optimized for low latency and high-volume workloads.

⚠ Deprecated✓ Public accessLLMMultimodalReasoning model

Context window

1M (1 048 576 tokenów)

tokens

Max output

65,536

tokens

Release date

17 June 2025

🏢Google DeepMindProducer 🏢GoogleOwner

Access:APIHostedDeployment:☁ Cloud

Overview

Gemini 2.5 Flash is a multimodal Google model from the Gemini 2.5 family, positioned as the best price-performance option for low-latency, high-volume tasks that require reasoning.

It is a hybrid reasoning model: it supports a configurable thinking budget that controls how much the model reasons before responding, balancing quality, cost and latency.

The model accepts text, image, audio and video as input and produces text output. Its context window is 1,048,576 input tokens, with a maximum of 65,536 output tokens. It supports function calling, structured output, context caching, code execution and grounding with Google Search.

Gemini 2.5 Flash is served through the Gemini API (Google AI Studio) and Vertex AI; the model weights are not publicly available. The gemini-2.5-flash version reached general availability (GA) on 17 June 2025.

Classification

LLMMultimodalReasoning model

Access & deployment

APIHosted

Cloud

Weights: Closed

Key parameters

📏 Context: 1M (1 048 576 tokenów)

✓ Tools · ✓ Fine-tuning

📥 Input: text, image, audio, video

Technical specification

Context window

1M (1 048 576 tokenów)

tokens

Max output tokens

65,536

tokens per response

License

Proprietary (Google)

Features:✓ Tool use✓ Fine-tuning

Modalities

⬇ Input

textimageaudiovideo

⬆ Output

textcode

Capabilities and applications

Native model capabilities

Advanced reasoning

The ability to perform multi-step, structured reasoning: analysing problems, planning steps, and drawing conclusions from hypotheses. Reasoning-first models (e.g. GPT-5.1 Thinking) dedicate a portion of inference to chains of thought before responding.

Category: reasoning

Multi-step reasoning

Carrying out multi-step chains of reasoning across long, complex tasks.

Category: reasoning

Extended thinking mode

A reasoning-model variant with a larger inference budget: more thinking cycles, higher answer precision at the cost of response time. Choice between 'standard' and 'extended' thinking is left to the user (e.g. the selector in GPT-5.2 Pro).

Category: reasoning

Coding

Generating, analysing and modifying code in many programming languages. Covers writing functions, debugging, refactoring, code review, and creating tests. Measured by benchmarks such as HumanEval and SWE-bench.

Category: coding

Agentic coding

Multi-hour, multi-step programming tasks performed autonomously by the model: cloning a repository, running tests, iterating on fixes, integrating with CLI tools. Characteristic of Codex variants (GPT-5.1-Codex-Mini, Codex-Max).

Category: coding

Long context

Support for large context windows — tens to hundreds of thousands (or millions) of input tokens. Enables analysis of entire codebases, long documents, and many parallel conversations without losing earlier information. GPT-5.1 supports 400,000 tokens.

Category: language

Multimodal understanding

Category: multimodal

Image understanding

Analysing and interpreting the content of images.

Category: vision

Audio understanding

Category: audio

Video understanding

The model's ability to analyse and interpret video content — recognising actions, motion, events and relationships between objects over time.

Category: video

Mathematical reasoning

The model's ability to solve mathematical tasks requiring multi-step reasoning — equations, proofs, combinatorics, geometry, calculus and competition-level problems.

Category: reasoning

Instruction following

Precisely following instructions contained in the prompt: response format, length, style, constraints (e.g. 'reply in six words'). GPT-5.1 significantly improved this capability compared to GPT-5.

Category: language

Multilingual

Competence in many natural languages (from a few to over a hundred): understanding, generation, translation, and code-switching within a single conversation. Frontier models support a wide range of languages with comparable quality.

Category: language

OCR

Recognising text within images and documents.

Category: vision

Structured output

Producing data in structured formats such as JSON.

Category: structured_generation

Tool use

The model's ability to call external functions, APIs and tools during a conversation: calculator, search engine, code editor, database. The model decides when and how to use a tool and interprets its result.

Category: planning

Function Calling

Category: planning

Chart understanding

Reading and interpreting charts, tables and diagrams.

Category: vision