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

Private AI

175 new models in the last 30 days

Every AI model. One place.

Text, image, video, 3D, audio and embedding models from every major lab. Compare them, see what everyone is using, and start with the best one for your task.

models
1,263
AI labs
126+
modalities
5
added this month
175
QwenOpenAIZ.AIAnthropicDeepSeekGoogleMistralMoonshot AIMiniMaxNVIDIAXiaomiMetaxAIAion LabsInclusionaiEva Unit 01TencentThinking MachinesUpstageAmazonLiquidNex AGISakanaHeabsyInceptionKitaniMoondreamPerceptronQwenOpenAIZ.AIAnthropicDeepSeekGoogleMistralMoonshot AIMiniMaxNVIDIAXiaomiMetaxAIAion LabsInclusionaiEva Unit 01TencentThinking MachinesUpstageAmazonLiquidNex AGISakanaHeabsyInceptionKitaniMoondreamPerceptron
Browse

What do you want to explore?

Text599 modelsChat, code, write and reason
Images240 modelsGenerate, edit and upscale
Videos214 modelsText and image to video
3D49 modelsTextured assets from a picture
Audio121 modelsSpeech, transcription, voices and music
Embeddings19 modelsAPI onlyVectors and reranking for search and RAGDecisions21 modelsAPI onlyTyped answers with probabilities for routing, classification and scoring
models
1,263
AI labs
126+
modalities
5
added this month
175
QwenOpenAIZ.AIAnthropicDeepSeekGoogleMistralMoonshot AIMiniMaxNVIDIAXiaomiMetaxAIAion LabsInclusionaiEva Unit 01TencentThinking MachinesUpstageAmazonLiquidNex AGISakanaHeabsyInceptionKitaniMoondreamPerceptron
Trending

Popular this week

The text models NanoGPT users run the most right now.

All 599 text models
  1. 1#1 Abliterated Model Large V2

    Abliteration.ai's default unrestricted large text reasoning model is derived from GLM-5.3 and weight-modified to reduce refusals compared with the base model. It supports native tool calling, structured output, automatic prompt caching, and a one-million-token context window.

  2. 2#2 Claude Opus 5.5

    Claude Opus 5.5 is Anthropic's flagship model for agentic coding, computer use, complex knowledge work, and long-running tasks, with improved efficiency and more natural communication.

  3. 3#3 GLM 5.3 Flash Uncensored

    GLM 5.3 Flash Uncensored is an uncensored fine-tune of the efficient 320B mixture-of-experts reasoning model with provider-dependent vision support, built for unrestricted chat, creative writing, coding, agentic work, tool use, and long-context tasks.

  1. 4GLM 5.3 Uncensored
  2. 5GLM 5.3
  3. 6GLM 5.3 Thinking
  4. 7Gemini 4 Argon
  5. 8Gemini 4 Carbon
Just landed

Recently added

The newest models across every type.

  • 3DNEW
    Hyper3D Low Poly

    Added today

  • 3DNEW
    Idol Max

    Added today

  • 3DNEW
    Idol Max Turbo

    Added today

  • TextNEW
    Tencent Hy4 Flash

    Added today

  • TextNEW
    DeepSeek V4.1 Flash 1017 Abliterated CSF (jcartu)

    Added yesterday

  • TextNEW
    GPT 6.2 Astra

    Added yesterday

  • TextNEW
    GPT 6.2 Astra Pro

    Added yesterday

  • TextNEW
    GPT 6.2 Sol

    Added yesterday

By the numbers

Leaderboards & benchmarks

What people actually use on NanoGPT, and how models score on independent benchmarks.

All benchmarks
84Benchmarks trackedFrom 8 independent sources
232Models with scoresCompared side by side
9.7%Usage share of the #1 modelAbliterated Model Large V2
66.0%Top overall scoreClaude Opus 5.5

Text model usage share

Weekly share of NanoGPT usage over the last 26 weeks

Full leaderboard
  1. 1. Abliterated Model Large V2

    9.7%

  2. 2. Claude Opus 5.5

    8.7%

  3. 3. GLM 5.3 Flash Uncensored

    7.8%

  4. 4. GLM 5.3 Uncensored

    7.6%

  5. 5. GLM 5.3

    5.5%

  6. 6. Gemini 4 Argon

    3.9%

  7. 7. Gemini 4 Carbon

    3.6%

  8. 8. Abliterated Large Flash

    3.5%

  9. 9. DeepSeek V4.1 Flash 1017 Uncensored

    3.5%

  10. 10. Kimi K3.1 Pro

    2.7%

Videos

Full ranking
  1. Wan 3.0 Reference-to-Video

    28.0%

  2. Seedance 2.5

    13.1%

  3. Seedance 2.5 Spicy

    11.1%

  4. Wan 3.0 Video Edit

    7.4%

  5. Wan 3.0 Prime

    5.7%

Images

Full ranking
  1. Seedream 5.0 Pro Edit

    23.3%

  2. Seedream 5.5

    13.5%

  3. Nano Banana 2.1

    7.5%

  4. GPT Image 2.5 Sunburst Edit

    6.6%

  5. Seedream 5.0 Pro

    6.3%

Text
Images
Video
#ModelScore$ / 1M tokens
  1. 1

    Claude Opus 5.5

    66.0%$8.00
  2. 2

    Gemini 4 Carbon

    60.2%$4.00
  3. 3

    Gemini 4 Argon

    50.1%$4.00
  4. 4

    Abliterated Model Large V2

    33.3%$3.50
  5. 5

    GPT 6.2 Astra

    33.3%$16.00
  6. 6

    Claude Fable 5.2

    31.8%$16.00
  7. 7

    Claude Fable 5.1

    30.0%$20.00
  8. 8

    Gemini 4 Neon

    27.0%–
  9. 9

    Claude Fable 5

    24.9%$20.00
  10. 10

    Muse Wonder

    24.7%$8.00

Source: NanoGPT Composite · LMArena data, CC BY 4.0 · Epoch AI data, CC BY 4.0 · Filtered to models on NanoGPT

Use cases

The best model for the job

Top picks per task, ranked by independent benchmarks and refreshed automatically.

All use cases

Write and fix code

“Refactor this module and add tests”

Gemini 4 CarbonTop pick ·89.2%
  1. 2Gemini 4 Neon70.6%
  2. 3Claude Opus 5.559.4%
Ranked by NanoGPT Composite

Creative writing

“Draft a short story in the style of a noir novel”

Claude Opus 5.5Top pick ·76.6%
  1. 2Claude Fable 5.172.4%
  2. 3Claude Opus 557.8%
Ranked by NanoGPT Composite

Roleplay & characters

“Stay in character as a ship captain”

Claude Opus 5Top pick ·96.9%
  1. 2Claude Opus 5.577.7%
  2. 3Kimi K358.5%
Ranked by NanoGPT Composite

Math & reasoning

“Prove this inequality step by step”

Gemini 4 ArgonTop pick ·100.0%
  1. 2Claude Fable 5.186.7%
  2. 3Claude Fable 574.9%
Ranked by NanoGPT Composite

Generate images

“A product shot on marble, soft morning light”

GPT Image 2.5 SunburstTop pick ·1428 Elo
  1. 2GPT Image 2.5 Flare1400
  2. 3Nano Banana 2.11328
Ranked by LMArena Dataset

Edit photos

“Swap the background for a beach at sunset”

GPT Image 2.5 SunburstTop pick ·1519 Elo
  1. 2GPT Image 2.5 Flare1484
  2. 3Grok Imagine Image 2.01437
Ranked by LMArena Dataset

Make videos

“A drone shot over a foggy forest at dawn”

Gemini Omni 1.2 FlashTop pick ·1540 Elo
  1. 2Gemini Omni Flash 1.11513
  2. 3Gemini Omni Flash1509
Ranked by LMArena Dataset

Animate an image

“Bring this portrait to life with a slow smile”

Gemini Omni 1.2 FlashTop pick ·1503 Elo
  1. 2Seedance 2.51494
  2. 3MiniMax H31493
Ranked by LMArena Dataset
Showcase

Made with models on NanoGPT

Sample outputs from the newest image and video models. Click one to try it.

Kling Image V4Try itIris-ZTry itIris-Z SmallTry itBRIA RMBG-3.0Try itMiniMax H3 ImageTry itSeedream 5.5Try itSeedream 5.5 Layer DecompositionTry itP-H3 ImageTry itSaffron lounge chair beneath a stone arch in a sunlit courtyardKling Image V4Try itSaffron lounge chair framed by a lavender colonnadeIris-ZTry itCobalt glass and blood orange on terracotta plinthsIris-Z SmallTry itTransparent cutout preserving the openwork of a crochet cardiganBRIA RMBG-3.0Try itMiniMax H3 Image — vinyl-inspired recording studio portraitMiniMax H3 ImageTry itSeedream 5.5 red ceramic teapot and lemon on a transparent backgroundSeedream 5.5Try itSeedream 5.5 isolated transparent teapot layerSeedream 5.5 Layer DecompositionTry itP-H3 Image cinematic portrait in a rainy taxiP-H3 ImageTry itKling Image V4Try itIris-ZTry itIris-Z SmallTry itBRIA RMBG-3.0Try itMiniMax H3 ImageTry itSeedream 5.5Try itSeedream 5.5 Layer DecompositionTry itP-H3 ImageTry it
3D to VideoTry itFLUX.1 Pro EraseTry itGPT Image 2.6 FlareTry itGPT Image 2.6 SunburstTry itHeyGen Avatar 6Try itHeyGen Avatar 6 TurboTry itLongCat Avatar 1.6Try itLTX-2.5 AlphaGenTry itLTX-2.5 HDRTry itMorgana v1.2Try itSeedance 2.5 FlashTry itTopaz Denoise MaxTry itTopaz Image UpscaleTry itTopaz Super FocusTry itTopaz Transparent UpscaleTry itVideo Background Remover ProTry itA rough 3D cabin layout restyled into a snowy pine forest3D to VideoTry itFLUX.1 Pro Erase removes a cup and small objects from a deskFLUX.1 Pro EraseTry itGPT Image 2.6 Flare markup edit: a cobalt-blue vase in a sunlit kitchenGPT Image 2.6 FlareTry itGPT Image 2.6 Sunburst paper-cut football celebration with blue jerseys and a CHAMPIONS signGPT Image 2.6 SunburstTry itHeyGen Avatar 6 stock-avatar speech exampleHeyGen Avatar 6Try itHeyGen Avatar 6 Turbo stock-avatar speech exampleHeyGen Avatar 6 TurboTry itLongCat Avatar 1.6 single-person talking exampleLongCat Avatar 1.6Try itLTX-2.5 AlphaGen — official paired processing exampleLTX-2.5 AlphaGenTry itLTX-2.5 HDR — SDR tone-mapped preview of the HLG masterLTX-2.5 HDRTry itA miniature observatory above clouds, generated with Morgana v1.2 (low quality)Morgana v1.2Try itA paper boat drifting across a quiet pond at sunriseSeedance 2.5 FlashTry itTopaz Denoise Max resultTopaz Denoise MaxTry itTopaz Standard V2 resultTopaz Image UpscaleTry itTopaz Super Focus V3 resultTopaz Super FocusTry itTopaz Transparent 4x resultTopaz Transparent UpscaleTry itSubject separated from the background with a transparent WebM alpha channelVideo Background Remover ProTry it3D to VideoTry itFLUX.1 Pro EraseTry itGPT Image 2.6 FlareTry itGPT Image 2.6 SunburstTry itHeyGen Avatar 6Try itHeyGen Avatar 6 TurboTry itLongCat Avatar 1.6Try itLTX-2.5 AlphaGenTry itLTX-2.5 HDRTry itMorgana v1.2Try itSeedance 2.5 FlashTry itTopaz Denoise MaxTry itTopaz Image UpscaleTry itTopaz Super FocusTry itTopaz Transparent UpscaleTry itVideo Background Remover ProTry it
Labs

Browse by creator

Model families from the labs that trained them.

Inference providers
  • Qwen

    70 models

  • OpenAI

    55 models

  • Z.AI

    42 models

  • Anthropic

    33 models

  • DeepSeek

    28 models

  • Google

    26 models

  • Mistral

    21 models

  • Moonshot AI

    14 models

  • MiniMax

    10 models

  • NVIDIA

    10 models

  • Xiaomi

    10 models

  • Meta

    9 models

  • xAI

    8 models

  • Aion Labs

    6 models

  • Inclusionai

    6 models

  • Eva Unit 01

    4 models

  • Tencent

    4 models

  • Thinking Machines

    4 models

  • Upstage

    4 models

  • Amazon

    3 models

  • Liquid

    3 models