Gemini 3.6 Flash
gemini-3.6-flash
Total Context
1M
Max Output
65.5K
Released
N/A
Input$1.5 / M tokens
Output$7.5 / M tokens
Gemini 3.1 Pro Preview
gemini-3.1-pro-previewGemini 3.1 Pro Preview is presented as a reliability-focused refinement of Gemini 3 Pro. Google’s preview notes emphasize better thinking, token efficiency, grounding, factuality, software engineering, agentic tool use, and multi-step execution. The preview label matters: it should be described as an advanced but evolving model for teams testing the next Pro behavior.
Context Window
1M tokens
Maximum Output
65.5K tokens
Release Date
Feb 19, 2026
Modalities
Gemini 3.1 Pro Preview Pricing
| Token Tier | Input Price | Output Price | Cache Read | Cache Create 5m |
|---|---|---|---|---|
| <=200K | $2/M | $12/M | $0.2/M | $0.375/M |
| >200K | $4/M | $18/M | $0.4/M | $0.375/M |
Gemini 3.1 Pro Preview API Capabilities
Reasoning
Supported
Tool calling
Supported
Temperature parameter
Supported
Attachments
Supported
Knowledge Base
2025-01-01
Endpoint Protocols
Completions APIMessages APIgemini
Gemini 3.1 Pro Preview Model Highlights
Gemini 3.1 Pro Preview combines improved reasoning, reliable multi-step execution and long-context multimodal analysis for demanding software and agent workflows.
Grounded Complex Reasoning
Reasons through difficult problems with improved factual consistency and more efficient use of reasoning tokens.
Reliable Agentic Execution
Plans and completes multi-step workflows with precise tool selection and consistent follow-through.
Long-Context Multimodality
Processes text, images, video, audio and PDFs within a context window exceeding one million input tokens.
Gemini 3.1 Pro Preview Use Cases
Gemini 3.1 Pro Preview is suited to complex software engineering, tool-driven agents and evidence-heavy multimodal research.
Complex Software Engineering
Analyze a large codebase, plan coordinated modifications and implement tested changes across multiple files.
Tool-Driven Agent Workflows
Choose and sequence tools across a multi-stage task, evaluate intermediate results and continue toward a defined outcome.
Multimodal Research Analysis
Review long documents, recordings and visual evidence, reconcile findings and produce a structured analytical report.
How to Use Gemini 3.1 Pro Preview via the TokenHub API
import OpenAI from "openai"
const client = new OpenAI({
apiKey: process.env.TOKENHUB_API_KEY,
baseURL: "https://us-api.tokenhub.com/v1",
})
const result = await client.chat.completions.create({})
console.log(result.choices[0]?.message?.content)Gemini 3.1 Pro Preview Benchmarks
Gemini 3.1 Pro Preview
| Index score | ||
|---|---|---|
| Artificial Analysis Intelligence Index | Artificial Analysis broad capability aggregate | 46.5 |
| Artificial Analysis Coding Index | Artificial Analysis software task aggregate | 55.5 |
| Knowledge & Reasoning | ||
| GPQA | Advanced science problem solving | 94.1% |
| HLE | Broad expert-level exam set | 44.7% |
| Coding & Engineering | ||
| SciCode | Scientific coding challenges | 58.9% |
| Terminal-Bench Hard | Hard terminal task execution | 53.8% |
| Instruction Following & Agent Tasks | ||
| IFBench | Prompt constraint adherence | 77.1% |
| AA-LCR | Long-context reasoning | 72.7% |
| τ²-Bench | Agent workflow tasks | 95.6% |
Metrics sourced from Artificial Analysis
Gemini 3.1 Pro Preview Comparisons
Frequently asked questions about Gemini 3.1 Pro Preview
Understand what Gemini 3.1 Pro Preview is, its best uses, distinguishing strengths, practical tradeoffs, and safe TokenHub integration guidance.
What kind of model is Gemini 3.1 Pro Preview?+
Gemini 3.1 Pro Preview is Google’s preview Pro model for advanced reasoning, software engineering, and reliable multi-step tool use. It is a preview model, so behavior, identifiers, and availability can change before stabilization.
What should teams use Gemini 3.1 Pro Preview for?+
Best-fit scenarios include complex multi-step reasoning, software engineering and code maintenance, and analysis of text and visual inputs. Test representative inputs and define measurable acceptance criteria before production.
Where does Gemini 3.1 Pro Preview have a clear technical advantage?+
Key strengths include strong reasoning on difficult problems, effective use of tools and function calls, and improved factual consistency and grounding. This combination is especially useful for software engineering and code maintenance.
When should a team choose another model instead of Gemini 3.1 Pro Preview?+
Consider another model when a stable interface and predictable behavior are mandatory, very low latency is the main requirement, or the workflow cannot include human review for important decisions. Verify important factual, legal, financial, medical, or operational outputs with qualified human review.
What should be checked before integrating Gemini 3.1 Pro Preview with TokenHub?+
In TokenHub, select the exact model identifier displayed for Gemini 3.1 Pro Preview, use the endpoint documented for your account, and authenticate with your TokenHub credentials. Pin the exact identifier shown by TokenHub, test behavior changes, and plan a migration path because preview aliases can be updated or withdrawn.
Ready to use Gemini 3.1 Pro Preview?
Use one API key to access Gemini 3.1 Pro Preview and more AI models through TokenHub.
Media and Discussions
Selected public videos and posts related to this model.
X (Twitter)
Reddit
YouTube