> ## Documentation Index > Fetch the complete documentation index at: https://docs.glif.app/llms.txt > Use this file to discover all available pages before exploring further. # Supported Models Glif has 120+ AI models for images, video, audio and text. You don't need to choose: Glif picks the right model for each step of your brief. If your team has a favorite, name it in your prompt or save it to memory. The full, current list is at [glif.app/models](https://glif.app/models). Models are added and retired often, so that page is always the source of truth. ## Intelligence tiers The intelligence tier is the agent that plans your work and chooses the tools. Higher tiers handle bigger briefs with more deliverables and cost more. | Tier | Cost | | :- | :- | | Lite | \$ | | Smart | \$\$ | | Genius | \$\$\$ (Pro plan and above) | Choose a tier in **Settings** or from the model menu in the composer. ## How Glif picks models Glif runs every model through the same set of standard tests and its team rates the results in blind head-to-head comparisons. The current default picks are shown at the top of [glif.app/models](https://glif.app/models), and you can see every test side by side at [glif.app/models/tests](https://glif.app/models/tests). Ratings are grouped by task, including **product photography**, **brand and text rendering**, **character consistency**, **text-to-image** and **image-to-video**. Each model's page shows where it ranks. ## Models for marketing work **Images** | Model | Good for | | :- | :- | | [GPT Image 2.5](https://glif.app/models/image/gpt-image-2-5) | Ads with text, product shots, transparent-background cutouts, infographics. Rated #1 for product photography and for brand and text rendering. | | [Nano Banana Pro](https://glif.app/models/image/nano-banana-pro) | Posters, infographics and composites from up to 14 reference images; text rendering; output up to 4K | | [Phota Image Editor](https://glif.app/models/image/phota-edit) | Editing from one or more reference images, up to 4K | | [Seedream 5.0 Pro](https://glif.app/models/image/seedream-5-pro), [MAI Image 2.5](https://glif.app/models/image/mai-image-25), [Krea 2](https://glif.app/models/image/krea-2) | Text-to-image and editing; Krea 2 accepts style references | | [P-Image Upscale](https://glif.app/models/image/p-image-upscale) | Upscaling for print and large displays, up to 128 MP | | [Bria Background Remover](https://glif.app/models/image/bria-background-remove) | Cutting products out of their background | **Video** | Model | Good for | | :- | :- | | [Seedance 2.5](https://glif.app/models/video/seedance-25) | Video with audio from text or an image; strong image-to-video ratings | | [Hailuo H3 Max](https://glif.app/models/video/hailuo-h3-max) | Video with synchronized audio, optional closing frame, 480p or 768p | | [Kling O3 Pro](https://glif.app/models/video/kling-o3-pro-image) | 3–15 second clips from an image, multi-shot prompts, optional audio, start and end frames | | [Seedance 2.0](https://glif.app/models/video/seedance-20) | Video with synchronized audio, start and end frames | | [Grok Imagine Video 1.5](https://glif.app/models/video/grok-imagine-video-15) | 1–15 second clips with audio from an image, 480p or 720p | | [Kling V3 4K](https://glif.app/models/video/kling-v3-4k-image) | 4K image-to-video | | [Motion Control (Kling V3)](https://glif.app/models/video/kling-v3-motion-control) | Copying motion onto a character | | [Wan 3.0](https://glif.app/models/video/wan-3-0) | Video up to 1080p | | [Edit Video (FLUX.3)](https://glif.app/models/video/edit-video-flux-3) | Restyling or editing an existing clip (up to 15 s) | | [Add Video Subtitles (VEED)](https://glif.app/models/video/video-subtitles-veed) | Captions on finished videos | | [Remove Video Background (VEED)](https://glif.app/models/video/video-remove-background-veed) | Cutting people or products out of video | | [Topaz Video Upscale](https://glif.app/models/video/topaz-video-upscale) | Upscaling finished video | **Audio** | Model | Good for | | :- | :- | | [ElevenLabs V4 Text to Speech](https://glif.app/models/audio/speech-elevenlabs-v3), [Seed Audio 1.0](https://glif.app/models/audio/speech-seed-audio), [Google Gemini TTS](https://glif.app/models/audio/speech-google-gemini) | Voiceover | | [Generate Music (ElevenLabs)](https://glif.app/models/audio/music-elevenlabs) | Background music from 10 to 300 seconds, or a structured composition plan | | [Generate Music (Lyria 3)](https://glif.app/models/audio/music-lyria-3) | Music clips (\~30 s) up to full songs with Lyria 3 Pro | | [Generate Music (MiniMax v2.6)](https://glif.app/models/audio/music-minimax-v2) | Songs with vocals, from your lyrics or written for you | | [Video to Sound Effects (Mirelo)](https://glif.app/models/audio/mirelo-sfx), [Generate Sound Effects (ElevenLabs)](https://glif.app/models/audio/sound-effects-elevenlabs) | Sound effects | | [Transcribe Audio (ElevenLabs Scribe)](https://glif.app/models/audio/transcribe-elevenlabs-scribe) | Transcripts | | [Split Audio (LALAL.AI)](https://glif.app/models/audio/split-audio-lalal) | Separating vocals from music | **Analysis** [Analyze Media (Gemini 3.1 Flash Lite)](https://glif.app/models/llm/gemini-31-flash-lite) reads images, video and audio, for example to describe or review an asset. ## Ask for a specific model Name the model in your prompt: "Make this with Nano Banana Pro." To always use a model for a kind of task, ask Glif to remember it: "Remember that we prefer Seedance 2.5 for UGC video." ## What models cost Each model's page lists its credit cost and what affects it, such as resolution or length. [glif.app/models/pricing](https://glif.app/models/pricing) compares them all. ## Voices and likeness Only use a voice or a person's likeness that you have the rights and consent to use. ## More on glif.app * [Models](https://glif.app/models) * [Model pricing](https://glif.app/models/pricing) * [Glif tests](https://glif.app/models/tests) * [Kling 3.0 Prompting Guide](https://glif.app/use-cases/kling-3-prompting-guide) * [AI Voiceover](https://glif.app/use-cases/ai-voiceover) *** ## Agent Reference ### Selection behavior * Glif chooses models per step. A goal-level brief works as well as one naming a model. * Name a model only when the user requests one. Use the exact name listed at `https://glif.app/models`. * Models marked **Deprecated** on glif.app/models should not be requested. * Treat the lists on this page as a snapshot; `https://glif.app/models` is the source of truth. ### Intelligence tiers | Tier | MCP `intelligence` value | | :- | :- | | Lite | `lite` | | Smart | `smart` | | Genius | `genius` (Pro plan and above) | * Per request: pass `intelligence` to `compose_project`. * Account default for new projects: `update_settings`. Read with `read_settings` or `whoami`. * If omitted, new projects use the account default; existing projects keep their tier. ### Model pages | Item | Format | | :- | :- | | Directory | `https://glif.app/models` (categories: `/models/image`, `/models/video`, `/models/audio`, `/models/llm`) | | Model page | `https://glif.app/models//` | | Pricing table | `https://glif.app/models/pricing` | | Tests | `https://glif.app/models/tests` | | Credit value | 1 credit = US\$0.01 | ### Credit costs (snapshot, from each model's page) | Model | Credits | | :- | :- | | GPT Image 2.5 | 0.7–73 per image (size, quality, references, prompt length) | | Nano Banana Pro | 18.09 (1K/2K) or 32.4 (4K) per image | | Phota Image Editor | 12.15 (1K) or 24.3 (4K) | | Seedream 5.0 Pro | 6.08 up to 1536×1536, 12.15 above | | Krea 2 | 4.05 (medium) or 8.1 (large) | | P-Image Upscale | 0.68 (≤4 MP) to 16.2 (≤128 MP) | | Hailuo H3 Max | 6.75/s (480p), 10.8/s (768p) | | Grok Imagine Video 1.5 | 10.8/s (480p), 18.9/s (720p), plus 1.35 per image input | | Kling O3 Pro | 15.12/s, or 18.9/s with audio | | Kling V3 4K | 56.7/s | | Motion Control (Kling V3) | 17.01/s (standard), 22.68/s (pro) | | Wan 3.0 | 6.75/s (480p), 13.5/s (720p), 27/s (1080p) | | Seedance 2.5 | 31–3,023 per clip, by resolution and duration | | Edit Video (FLUX.3) | 4.05/s at 720p | | Add Video Subtitles (VEED) | 13.5/min (basic preset) | | ElevenLabs V4 TTS | 10.8 per 1,000 characters | | Seed Audio 1.0 TTS | 25.31 per minute | | Generate Music (ElevenLabs) | 20.25 per minute | | Generate Music (MiniMax v2.6) | 20 per track | | Video to Sound Effects (Mirelo) | 1.49 per second | Costs change; confirm on the model page before quoting a user an estimate. Agent (intelligence tier) costs are additional. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.