Private AI

Nano GPT logo
NanoGPT
Nano GPT logo
NanoGPT
Nano GPT logo
NanoGPT
Search
 
 
 
 
All conversations

Create

Chat
Studio

Explore

Models
Pricing
AI Apps
Blog

Library

Gallery
Workspace
Usage & billing
API
Suggestions and bugs
Use cases

|

Terms

|

Privacy

|

Refunds
Usage & billing
Teams
Referrals
Presets
Settings
FAQPress & mediaBrand assetsBenchmarksNano RPCEarn crypto
  1. Models
  2. Text
  3. Unsloth

Gemma 3 12B IT

Gemma 3 has a large, 128K context window, multilingual support in over 140 languages, and is available in more sizes than previous versions.

  • Native PDF input
Model weights

Pricing

Auto routing · per 1M tokens
Input
$0.272
Output
$0.272
Cache read
$0.136
Compare provider prices

Specifications

Context window
128K
Max output
131.1K
Parameters
12B

Benchmarks

Sourced from LMArena.

Arena Score

1341.7

Overall Rank

#238 / 421

Votes

3,829

Confidence Interval

1332.2 - 1351.3

Category Scores

Coding

#300 / 416

543 votes

1315.8

Math

#234 / 404

389 votes

1317.7

Longer Query

#229 / 399

371 votes

1346.6

Creative Writing

#206 / 419

600 votes

1332.5

Instruction Following

#241 / 421

1,145 votes

1321.0

Hard Prompts

#258 / 421

977 votes

1331.9

Additional Categories
16

German

#167 / 320

166 votes

1371.8

Industry Business And Management And Financial Operations

#183 / 414

364 votes

1386.9

Industry Legal And Government

#198 / 393

217 votes

1383.1

Russian

#201 / 385

285 votes

1357.3

Industry Life And Physical And Social Science

#219 / 419

760 votes

1369.0

Non English

#219 / 421

1,499 votes

1335.5

Multi Turn

#222 / 419

471 votes

1344.3

Industry Mathematical

#225 / 401

333 votes

1349.3

Industry Entertainment And Sports And Media

#230 / 419

697 votes

1305.5

Exclude Ties

#238 / 421

2,526 votes

1295.7

Industry Writing And Literature And Language

#242 / 420

1,010 votes

1313.0

English

#249 / 421

2,330 votes

1349.8

Hard Prompts English

#262 / 419

700 votes

1338.7

Industry Medicine And Healthcare

#269 / 390

155 votes

1326.2

Industry Software And It Services

#285 / 421

867 votes

1334.1

Expert

#290 / 371

186 votes

1274.0

Published 2026-11-09 · Matched as Gemma 3 12B IT · CC BY 4.0

LMArena Dataset

Leaderboard results

BenchmarkScoreRank on NanoGPT
Epoch AI Benchmarking Hub

Epoch AI data, CC BY 4.0 · Updated Nov 12, 2026

GPQA Diamond

39.5%

#116 of 119

OTIS Mock AIME 2024–2025

16.7%

#111 of 121

EQ-Bench

EQ-Bench data, MIT · Updated Nov 12, 2026

Creative Writing v3

1131

#82 of 96

Longform Writing

41.8

#79 of 97

UGI Leaderboard

UGI Leaderboard data, Apache-2.0 · Updated Oct 16, 2026

UGI Score

20.9

#127 of 149

UGI Willingness

3.0

#77 of 149

UGI Writing

29.9

#126 of 149

Filtered and ranked to models on NanoGPT.

Providers

Choose or restrict providers with the API. Provider-selection prices include the selection markup. Provider selection guide · Pricing API (JSON, USD per 1K tokens)

Auto
Auto
Recommended
$0.272$0.272$0.136N/AN/A
Kitani
Kitani
INT4
GLB
$0.265$0.265$0.132N/AN/A
Privacy:ZDR, prompts not retained
Batch API
ProviderInputOutputCached input
Auto
Autounsloth/gemma-3-12b-it
$0.272/M$0.272/M$0.136/M

Batch rows for this model use normal routing and cost the same as standard API requests. Set provider in a row to choose a provider.

Related text models

Compare Gemma 3 12B IT with similar models from the same provider or model family.

Gemma 3 27B IT

unsloth/gemma-3-27b-it

Gemma 3 has a large, 128K context window, multilingual support in over 140 languages, and is available in more sizes than previous versions.

Gemma 3 4B IT

unsloth/gemma-3-4b-it

Gemma 3 has a large, 128K context window, multilingual support in over 140 languages, and is available in more sizes than previous versions.

Gemini 4 Carbon

google/gemini-4-carbon

Google's Gemini 4 Carbon is a reasoning model for complex coding, scientific and mathematical problem solving, and multi-step tasks that use tools. It accepts text, images, audio, video, and PDFs and produces text, with tool calling and structured outputs. Its 1,048,576-token context is shared across input, thinking, and output. Responses support up to 32,768 combined thinking and output tokens, with low, medium, and high thinking levels; medium is the default.

Gemini 4 Argon

google/gemini-4-argon

Google's Gemini 4 Argon is a flagship reasoning model for complex software engineering, mathematical and scientific problem solving, and multi-step agents. It accepts text, images, audio, video, and PDFs, with tool calling and structured outputs. Its 1,048,576-token context is shared across input, thinking, and output; synchronous output is capped at 65,536 tokens.

Gemini 3.8 Flash

google/gemini-3.8-flash

Google's fast multimodal model for agentic workloads, including coding, tool use, image understanding, PDF and document extraction, audio, and video. Its capabilities, limits, reasoning behavior, and pricing currently mirror Gemini 3.7 Flash.

Gemini 3.7 Flash

google/gemini-3.7-flash

Google's frontier-performance Flash model for multimodal and agentic workloads, including coding, tool use, image understanding, PDF and document extraction, audio, and video. Google reports 65.3% on DeepSWE v1.1, up from 49.0% for Gemini 3.6 Flash, and 34% on GDP.pdf, up from 14%.