Skip to content
llmwaves

Gemini 3.1 Pro Preview

Google · released Feb 19, 2026

ProprietaryReasoningTool useStructured outputVisionAudio inVideo in

Gemini 3.1 Pro Preview is Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across complex workflows. Building on the multimodal foundation...

Specification

Context window
1.05M
Max output
66K
Knowledge cutoff
Not stated
Parameters
Undisclosed
Licence
Proprietary
Serving providers
2
Moderated
No
Uptime
100.0%

Intelligence

29.7

82th percentile

Coding

68.8

Coding Index

Agentic

8.2

Agentic Index

Output speed

90 t/s

Median across providers

Latency

2.59s

Time to first token

Cost per task

$1.06

Estimated

Benchmarks

Where the score comes from

The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.

Evaluation scores

Percentage correct · higher is better

  • τ²-bench (Telecom)
    95.6%
  • GPQA Diamond
    94.1%
  • AA-LCR (long context)
    82.0%
  • IFBench
    77.1%
  • SciCode
    58.7%
  • Terminal-Bench Hard
    53.8%
  • Humanity's Last Exam
    47.0%

An evaluation missing from this list was not run for this model — it is not a zero.

View as table
EvaluationScore
τ²-bench (Telecom)95.6%
GPQA Diamond94.1%
AA-LCR (long context)82.0%
IFBench77.1%
SciCode58.7%
Terminal-Bench Hard53.8%
Humanity's Last Exam47.0%

Against its peers

Intelligence Index · this model highlighted, nearest peers in grey

Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.

View as table
ModelIntelligence
Ring-2.6-1T31.1
GPT-5.230.4
Gemini 3.1 Pro Preview29.7
Qwen3.7 Max29.5
MiniMax M329.2
GPT-5.2-Codex28.5
Solar Pro 428.2
GLM 527.9

Percentile among all indexed models

Intelligence82th
Coding78th
Agentic29th
Terminal work97th
Arena Elo87th

Pricing

What it costs to run

List prices per million tokens, plus what one representative task works out to.

List price

Input / 1M tokens
$2
Output / 1M tokens
$12
Cached input / 1M
$0.2
Blended 3:1
$4.5

One task, estimated

$1.06

Input tokens
50,000
Output tokens
80,000
Profile
Reasoning

Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).

Serving providers

Speed and latency figures are medians across these providers, so a widely-served model reports a blend rather than any single endpoint.

Arena

Head-to-head generation quality

Elo from pairwise judgements, broken out by the kind of thing the model was asked to build.

Overall Elo
1,283
Win rate
67.8%
Strongest at
SVG
Tournaments
4,620

Elo by category

Dot position on a 1,050–1,300 scale · Elo has no meaningful zero

Agent categories (full-stack apps, mobile apps) are only scored for models tested in agent harnesses.

View as table
CategoryEloWin rate
SVG128367.8%
Websites123653.8%
3D scenes123558.4%
UI components122854.0%
Data visualisation121052.1%
Game development120552.1%
Mobile apps109143.2%
Full-stack apps106840.8%