Current defaults
Why did we select these models?
Models
89 video, image and audio models, tested on the same prompts and rated in blind votes by the Glif team and an AI judge. The Glif agent picks the right one for each job.
Model leaderboard
Top text-to-image models. Score out of 100: how often people would pick this model's output over an average model's on the same tests. 50 is average.
Scores come from 2,042 blind votes by 10 Glif team members, plus 1,279 AI judge scores, each counting as half a vote.
See the boardExample test
Product placement
Preserves the reference can exactly (label, colors, proportions). Composes a believable bench scene in Ghent. The can should be sized proportionally.
Winner: Seedream V4.5, picked from 23 models by an AI judge.
See the results
FLUX.2 Pro
#2 of 13
FLUX 3 Image
#3 of 13
GPT Image 2.5
#3 of 13
Video Generation
Models for video generation · 69 models
Glif default
Hailuo H3 Max (MiniMax)
MiniMax
Detailed clips with synchronized sound, guided by text, frames or mixed media references.
Glif default
Grok Imagine Video 1.5
xAI
Animate a photo into a 480p or 720p clip with sound, up to 15 seconds long.
Glif default
Seedance 2.0
ByteDance
Create up to 15 seconds of 1080p video with synchronized speech and camera control.
Glif default
Kling O3 Pro · Image to video
Kuaishou Technology
Turn a starting image into a 3–15 second 1080p clip with synchronized sound.
New
Kling 4.0
Kuaishou Technology
Create 3–30 second clips from text or a starting image, with optional end frames and multiple shots.
New
FLUX.3 Max
Black Forest Labs
Create higher-quality 5–20 second clips with sound from text, a starting image or start and end frames.
Image Generation
Models for creating images · 38 models
Glif default
Krea 3
Krea
Glif default
Nano Banana 2.1
Google DeepMind
Fast images that keep characters consistent across up to 14 references.
Glif default
Ideogram V4.5
Ideogram
Accurate text and typography for logos, posters, packaging and signs.
Glif default
GPT Image 2.5
OpenAI
Precise image edits, readable text and transparent backgrounds.
Glif default
Muse Image
Meta
Precise reference-based edits and fine details such as text, charts and QR codes.
Glif default
Gemini 3.1 Flash
Google DeepMind
Fast image generation and editing with readable text and up to 4K output.
Audio
Models for speech, music, transcription and audio separation · 18 models
Glif default
ElevenLabs · Generate music
ElevenLabs
Songs and instrumentals from a prompt or a detailed plan with sections and lyrics.
Mirelo · Video to sound effects
Mirelo
Seed Audio 1.0 Text to Speech
ByteDance
Multilingual speech with 20 preset voices or voice cloning from audio samples.
Lyria 3 · Generate music
Google DeepMind
Short clips or full songs with control over mood, instruments, structure and vocals.
MiniMax v2.6 · Generate music
MiniMax
Complete songs with singing, backing music and arrangements from your lyrics.
LALAL.AI · Split audio
LALAL.AI