# varg.ai models

> Every generative model on varg — video, image, speech, music, transcription, editing — behind one REST API (https://api.varg.ai/v2) and one MCP server (https://mcp.varg.ai/mcp) on one credit balance. Each link is that model's full spec: inputs, limits, provider options, routes and prices.

How to call any model:
- REST: `POST https://api.varg.ai/v2/tools/{tool}/call` with `Authorization: Bearer $VARG_API_KEY` and a JSON body that includes `"model"`. Returns 202 with a job; poll `GET /v2/jobs/{id}`.
- MCP: `call_tool({ tool, input })` — `input` is the same JSON body.
- Schema: `GET https://api.varg.ai/v2/tools/{tool}?model={model}` (public JSON Schema). Price first: `POST https://api.varg.ai/v2/estimate`.

Generated from the live catalog (version 2026-10-16T04:35:24.302666+00:00). Every spec in one file: https://varg.ai/models/llms-full.txt

## Video (tool: video)

- [ByteDance Video Upscaler](https://varg.ai/models/bytedance_upscaler.md): `bytedance_upscaler` — ByteDance video upscaler. Affordable video enhancement. From 27 credits.
- [Creatify Boreal](https://varg.ai/models/creatify_boreal.md): `creatify_boreal` — Creatify Boreal. Product, UGC and presenter videos with synchronized native speech and sound from a structured prompt, optionally animating an image or driven by your audio; 1-20 seconds, 720p to 2K. From 2 credits.
- [FLUX 3](https://varg.ai/models/flux_3.md): `flux_3` — Black Forest Labs FLUX 3. Latest flagship video model — text-to-video, image-to-video, and video-to-video with native audio. 720p/1080p, 5-10s. From 90 credits.
- [FLUX 3 Draft](https://varg.ai/models/flux_3_draft.md): `flux_3_draft` — Black Forest Labs FLUX 3 Draft. Fast ~1/3 cost preview mode for FLUX 3 — text-to-video, image-to-video, video-to-video. HD only, 5-10s. From 32 credits.
- [Grok Imagine Video](https://varg.ai/models/grok_imagine.md): `grok_imagine` — xAI Grok video generation. Text-to-video and image-to-video. From 105 credits.
- [Grok Imagine 1.5](https://varg.ai/models/grok_imagine_1_5.md): `grok_imagine_1_5` — Grok Imagine Video 1.5 by xAI. Text, image and multi-reference video with native audio, 1-15 seconds up to 1080p. From 9 credits.
- [Grok Imagine 1.5 Lite](https://varg.ai/models/grok_imagine_1_5_lite.md): `grok_imagine_1_5_lite` — Grok Imagine Video 1.5 Lite by xAI. The cheapest video with native audio: text or image to video, 1-15 seconds up to 1080p. From 3 credits.
- [Grok Imagine Video Edit](https://varg.ai/models/grok_imagine_edit.md): `grok_imagine_edit` — xAI Grok video editing. Transform existing videos with text prompts. From 105 credits.
- [Happy Horse](https://varg.ai/models/happy_horse.md): `happy_horse` — Alibaba video generation. Expressive character motion. From 45 credits.
- [HeyGen Avatar](https://varg.ai/models/heygen_avatar.md): `heygen_avatar` — HeyGen AI avatar video. Generate talking avatar videos from text scripts. From 1050 credits.
- [HeyGen Avatar IV](https://varg.ai/models/heygen_avatar_iv.md): `heygen_avatar_iv` — HeyGen Interactive Avatar. Next-gen avatar with improved visual fidelity. From 1050 credits.
- [Heygen Avatar V](https://varg.ai/models/heygen_avatar_v.md): `heygen_avatar_v` From 1050 credits.
- [heygen_image_video](https://varg.ai/models/heygen_image_video.md): `heygen_image_video` From 1050 credits.
- [HeyGen Video 1](https://varg.ai/models/heygen_video_1.md): `heygen_video_1` — HeyGen Video 1. Text, image and multi-reference video with generated sound, 5-15 seconds up to 2K. From 11 credits.
- [ID-V2V](https://varg.ai/models/id_v2v.md): `id_v2v` — ID-V2V (Eyeline Labs / Netflix research). Restyle a video's scene, lighting and look from an edited first frame while keeping the subjects' identity, expressions, gaze and motion; 480p or 720p, 81 frames (about 5 s) by default, up to 241 via provider_options.fal.num_frames. From 21 credits.
- [ID-V2V Relight](https://varg.ai/models/id_v2v_relight.md): `id_v2v_relight` — ID-V2V Relight. Change a video's lighting from a relit first frame, keeping the scene, subjects and the original performance; 480p or 720p, 81 frames (about 5 s) by default, up to 241 via provider_options.fal.num_frames. From 21 credits.
- [Kling v2](https://varg.ai/models/kling_v2.md): `kling_v2` — Master tier. Stable and well-tested. From 59 credits.
- [Kling v2.5](https://varg.ai/models/kling_v2_5.md): `kling_v2_5` — Turbo variant. Fast generation with good quality. From 59 credits.
- [Kling v2.6](https://varg.ai/models/kling_v2_6.md): `kling_v2_6` — Pro tier with improved motion consistency. From 59 credits.
- [Kling v2.6 Motion](https://varg.ai/models/kling_v2_6_motion.md): `kling_v2_6_motion` — Motion control variant. Precise camera and subject movement. From 59 credits.
- [Kling v3](https://varg.ai/models/kling_v3.md): `kling_v3` — Latest Kling model. Pro tier. Text-to-video and image-to-video. From 59 credits.
- [Kling v3 4k Image To Video](https://varg.ai/models/kling_v3_4k_image_to_video.md): `kling_v3_4k_image_to_video` From 133 credits.
- [Kling v3 4k Reference To Video](https://varg.ai/models/kling_v3_4k_reference_to_video.md): `kling_v3_4k_reference_to_video` From 133 credits.
- [Kling v3 Pro Motion Control](https://varg.ai/models/kling_v3_pro_motion_control.md): `kling_v3_pro_motion_control` From 530 credits.
- [Kling v3 Pro Reference To Video](https://varg.ai/models/kling_v3_pro_reference_to_video.md): `kling_v3_pro_reference_to_video` From 59 credits.
- [Kling v3 Standard](https://varg.ai/models/kling_v3_standard.md): `kling_v3_standard` — Kling O3 standard tier. Good balance of quality and cost. From 59 credits.
- [Kling v3 Standard Motion Control](https://varg.ai/models/kling_v3_standard_motion_control.md): `kling_v3_standard_motion_control` From 397 credits.
- [Kling v3 Standard V2V Reference](https://varg.ai/models/kling_v3_standard_v2v_reference.md): `kling_v3_standard_v2v_reference` From 67 credits.
- [Lipsync](https://varg.ai/models/lipsync.md): `lipsync` — Basic lip-sync model. Affordable talking head generation. From 53 credits.
- [LTX-2](https://varg.ai/models/ltx_2_19b_distilled.md): `ltx_2_19b_distilled` — Lightricks LTX-2 distilled. Cheapest video model. From 53 credits.
- [LTX-2 Audio-to-Video](https://varg.ai/models/ltx_2_a2v.md): `ltx_2_a2v` — LTX-2 audio-driven video generation. Create videos driven by audio input. From 6 credits.
- [Minimax H3](https://varg.ai/models/minimax_h3.md): `minimax_h3` — MiniMax H3. Frontier 2K video model with native stereo audio. Text, image, or multimodal reference inputs. From 27 credits.
- [MiniMax H3 LoRA](https://varg.ai/models/minimax_h3_lora.md): `minimax_h3_lora` — MiniMax H3 LoRA. H3 with your own LoRA weights (up to 3, URL or Hugging Face repo): text, image or references to video, 480p to 4K, 5-15 seconds. From 34 credits.
- [MiniMax H3 Max](https://varg.ai/models/minimax_h3_max.md): `minimax_h3_max` — MiniMax H3 Max. fal's post-trained H3 with stronger prompt adherence and aesthetics: text, first/last frame or multimodal references, 480p-1080p, up to 15 seconds; can sync to an audio track. From 6 credits.
- [MiniMax H3 Max 3D to Video](https://varg.ai/models/minimax_h3_max_3d_to_video.md): `minimax_h3_max_3d_to_video` — MiniMax H3 Max 3D to Video. Turn a Blender render or 3D previs (up to 15 s) into photoreal video with the same layout, camera and timing; reference images set the look. 480p-1080p. From 95 credits.
- [MiniMax H3 Max Camera Controls](https://varg.ai/models/minimax_h3_max_camera_controls.md): `minimax_h3_max_camera_controls` — MiniMax H3 Max Camera Controls. Move the camera around a single image with keyframed azimuth, elevation and distance while the scene stays put. Up to 15 s, 480p-1080p. From 6 credits.
- [MiniMax H3 Max Extend Video](https://varg.ai/models/minimax_h3_max_extend.md): `minimax_h3_max_extend` — MiniMax H3 Max Extend Video. Continue a video with up to 15 s of prompt-guided footage; returns the full video or only the new part. 480p to 2K. From 5 credits.
- [MiniMax H3 Max Insert Video](https://varg.ai/models/minimax_h3_max_insert.md): `minimax_h3_max_insert` — MiniMax H3 Max Insert Video. Generate a new 5-13 s scene between two points of an existing video from a prompt or references, then cut back to the original footage. From 27 credits.
- [MiniMax H3 Max Lip Sync](https://varg.ai/models/minimax_h3_max_lipsync.md): `minimax_h3_max_lipsync` — MiniMax H3 Max Lip Sync. Talking video from one image and an audio track (5 s to 15 min); the video follows the audio length. 480p to 2K. From 32 credits.
- [MiniMax H3 Max Recast](https://varg.ai/models/minimax_h3_max_recast.md): `minimax_h3_max_recast` — MiniMax H3 Max Recast. Swap the people in a 5-30 s video for people from up to four photos, keeping motion, gestures, camera, cuts and soundtrack. 768p or 1080p. From 158 credits.
- [MiniMax H3 Max Relight](https://varg.ai/models/minimax_h3_max_relight.md): `minimax_h3_max_relight` — MiniMax H3 Max Relight. Relight a video (up to 15 s) from a lighting-sphere image: light color, direction and softness change while subjects, motion, camera and audio stay. 480p to 2K. From 13 credits.
- [MiniMax H3 Max · 16-bit Pixel](https://varg.ai/models/minimax_h3_max_style_16bit_pixel.md): `minimax_h3_max_style_16bit_pixel` — MiniMax H3 Max · 16-bit Pixel. 16-bit pixel-art video with audio from a prompt or a first-frame image, 768p, 5-15 seconds. From 42 credits.
- [MiniMax H3 Max · Hand Drawn](https://varg.ai/models/minimax_h3_max_style_hand_drawn.md): `minimax_h3_max_style_hand_drawn` — MiniMax H3 Max · Hand Drawn. Hand-drawn pencil animation video with audio from a prompt or a first-frame image, 768p, 5-15 seconds. From 42 credits.
- [MiniMax H3 Max · Low Poly](https://varg.ai/models/minimax_h3_max_style_low_poly.md): `minimax_h3_max_style_low_poly` — MiniMax H3 Max · Low Poly. Retro low-poly 3D video with audio from a prompt or a first-frame image, 768p, 5-15 seconds. From 42 credits.
- [MiniMax H3 Max · Retro Toon 70s](https://varg.ai/models/minimax_h3_max_style_retro_toon_70s.md): `minimax_h3_max_style_retro_toon_70s` — MiniMax H3 Max · Retro Toon 70s. 1970s hand-painted cartoon video with audio from a prompt or a first-frame image, 768p, 5-15 seconds. From 42 credits.
- [MiniMax H3 Max · VHS](https://varg.ai/models/minimax_h3_max_style_vhs.md): `minimax_h3_max_style_vhs` — MiniMax H3 Max · VHS. VHS-tape video with audio from a prompt or a first-frame image, 768p, 5-15 seconds. From 42 credits.
- [MiniMax H3 Max Turbo](https://varg.ai/models/minimax_h3_max_turbo.md): `minimax_h3_max_turbo` — MiniMax H3 Max Turbo. Nearly H3 Max quality at half the price and a fraction of the wait, up to 1080p and 15 seconds; can sync to an audio track. From 4 credits.
- [MiniMax H3 Max Turbo Extend Video](https://varg.ai/models/minimax_h3_max_turbo_extend.md): `minimax_h3_max_turbo_extend` — MiniMax H3 Max Turbo Extend Video. Half-price Turbo version of Extend Video: up to 15 s of new footage after an existing clip, 480p to 2K. From 3 credits.
- [Omni Flash](https://varg.ai/models/omni_flash.md): `omni_flash` From 4 credits.
- [Gemini Omni Flash 1.1](https://varg.ai/models/omni_flash_1_1.md): `omni_flash_1_1` — Gemini Omni Flash 1.1 by Google. Fast video with native audio from 720p to 4K: text, first/last frame, multi-reference and video editing. From 4 credits.
- [OmniHuman v1.5](https://varg.ai/models/omnihuman_v1_5.md): `omnihuman_v1_5` — ByteDance OmniHuman. Full-body animation from image + audio. From 1008 credits.
- [omnihuman_v2](https://varg.ai/models/omnihuman_v2.md): `omnihuman_v2` From 2205 credits.
- [Pixelcut Looping Video](https://varg.ai/models/pixelcut_looping_video.md): `pixelcut_looping_video` — Pixelcut Looping Video. Turn one product photo into a seamless 5-15 second loop with a locked camera: subtle ambient motion or a full 360° spin; 480p to 1080p. From 42 credits.
- [PixVerse VibeMV](https://varg.ai/models/pixverse_vibemv.md): `pixverse_vibemv` — PixVerse VibeMV. Music video for a whole song (10 s to 6 min) in one of 15 visual styles or your own style image, with an optional character reference and lyric subtitles; 720p or 1080p. From 63 credits.
- [Seedance 2](https://varg.ai/models/seedance_2.md): `seedance_2` — ByteDance Seedance 2. High-quality video from text, images, or video editing. From 30 credits.
- [Seedance 2.5](https://varg.ai/models/seedance_2_5.md): `seedance_2_5` — ByteDance Seedance 2.5 — the flagship Seedance model. High-quality video from text, with optional image (≤9), video (≤3), and audio (≤3) references. 480p/720p, 4-30s. From 62 credits.
- [Seedance 2.0 Fast](https://varg.ai/models/seedance_2_fast_preview.md): `seedance_2_fast_preview` — ByteDance Seedance 2 Fast. Lower cost variant of Seedance 2. Supports 5, 10, or 15 second duration. From 42 credits.
- [Seedance 2 Less Restriction](https://varg.ai/models/seedance_2_less_restriction.md): `seedance_2_less_restriction` From 33 credits.
- [Seedance 2 Mini](https://varg.ai/models/seedance_2_mini.md): `seedance_2_mini` — ByteDance Seedance 2.0 Mini via PiAPI — strict moderation, cheapest Seedance tier. Duration 4-15 seconds, 720p only. From 22 credits.
- [Seedance 2 Mini Less Restriction](https://varg.ai/models/seedance_2_mini_less_restriction.md): `seedance_2_mini_less_restriction` — ByteDance Seedance 2.0 Mini via PiAPI — relaxed general content review, explicit opt-in only. Content involving minors, real-person identity or faces, graphic violence and hate remains blocked. Duration 4-15 seconds. From 33 credits.
- [Seedance 2.0](https://varg.ai/models/seedance_2_preview.md): `seedance_2_preview` — ByteDance Seedance 2. High-quality video from text, images, or video editing. Supports 5, 10, or 15 second duration. From 53 credits.
- [SeedVR Video Upscale](https://varg.ai/models/seedvr_video.md): `seedvr_video` — ByteDance SeedVR video upscaler. Good quality video enhancement. From 32 credits.
- [Sima Video Upscaler](https://varg.ai/models/sima_video_upscaler.md): `sima_video_upscaler` — Sima video upscaler. Most affordable video upscaling. From 16 credits.
- [Sync v2](https://varg.ai/models/sync_v2.md): `sync_v2` — Sync Labs v2. Good lip-sync at lower cost. From 53 credits.
- [Sync v2 Pro](https://varg.ai/models/sync_v2_pro.md): `sync_v2_pro` — Sync Labs v2 Pro. Best lip-sync quality. Drives talking head videos. From 84 credits.
- [Sync v3](https://varg.ai/models/sync_v3.md): `sync_v3` — Sync Labs v3. Latest lip-sync model with improved quality. From 105 credits.
- [Topaz Video Upscale](https://varg.ai/models/topaz_video.md): `topaz_video` — Topaz-quality video upscaling. Best video enhancement quality. From 53 credits.
- [VEED Fabric](https://varg.ai/models/veed_fabric_1_0.md): `veed_fabric_1_0` — VEED Fabric. Image + audio to talking video. From 473 credits.
- [Veo 3.1](https://varg.ai/models/veo_3_1.md): `veo_3_1` From 84 credits.
- [Veo 3 1 Fast](https://varg.ai/models/veo_3_1_fast.md): `veo_3_1_fast` From 42 credits.
- [Veo 3 1 Lite](https://varg.ai/models/veo_3_1_lite.md): `veo_3_1_lite` From 13 credits.
- [Vidu Q4](https://varg.ai/models/vidu_q4.md): `vidu_q4` — Vidu Q4. Video with native audio from a first frame, or from reference images and voice clips; 3-16 seconds, 540p to 4K. No text-only mode. From 15 credits.
- [Wan 2.5](https://varg.ai/models/wan_2_5.md): `wan_2_5` — Alibaba Wan 2.5. Strong text-to-video at competitive price. From 27 credits.
- [Wan 2.5 Preview](https://varg.ai/models/wan_2_5_preview.md): `wan_2_5_preview` — Preview variant with lip-sync support. Cheapest Wan option. From 27 credits.
- [Wan 3.0](https://varg.ai/models/wan_3.md): `wan_3` — Alibaba Wan 3.0. Top-ranked text-to-video — up to 30s at 1080p with native audio, first/last-frame animation, and references (≤10 images, ≤5 videos, ≤5 audio). 480p/720p/1080p, 2-30s. From 11 credits.

## Images (tool: image)

- [Aura SR](https://varg.ai/models/aura_sr.md): `aura_sr` — Aura SR image upscaler. Free on provider, minimal cost. From 0 credits.
- [Bria FIBO-Edit 1.5 Expand](https://varg.ai/models/bria_fibo_edit_1_5_expand.md): `bria_fibo_edit_1_5_expand` — Bria FIBO-Edit 1.5 outpainting: extends one image to one of 9 aspect ratios or a custom canvas up to 25 MP, with an optional prompt for the new area. Original pixels are kept at full resolution. $0.04 per image. From 5 credits.
- [Bria Product Holding (FIBO-Edit 1.5)](https://varg.ai/models/bria_fibo_edit_1_5_product_holding.md): `bria_fibo_edit_1_5_product_holding` — Bria Product Holding on FIBO-Edit 1.5. Put the product from 1-3 reference images into the hand of the person in the first image; only the arm and hand move. From 5 credits.
- [Bria Virtual Try-On (FIBO-Edit 1.5)](https://varg.ai/models/bria_fibo_edit_1_5_virtual_try_on.md): `bria_fibo_edit_1_5_virtual_try_on` — Bria Virtual Try-On on FIBO-Edit 1.5. Dress the person in the first image in 1-3 garments or accessories from the other images, keeping identity, pose and background. From 5 credits.
- [Bria Guided Remove Background](https://varg.ai/models/bria_guided_remove_background.md): `bria_guided_remove_background` — Background removal from one image, optionally steered by a prompt that adds, drops or isolates objects ('only the chair'). Keeps the original pixels and returns a real alpha channel. Flat $0.03 per image. From 4 credits.
- [CCSR](https://varg.ai/models/ccsr.md): `ccsr` — CCSR image upscaler. Free on provider, minimal cost. From 0 credits.
- [chatgpt_image_latest](https://varg.ai/models/chatgpt_image_latest.md): `chatgpt_image_latest` From 105 credits.
- [Clarity Upscaler](https://varg.ai/models/clarity_upscaler.md): `clarity_upscaler` — Clarity upscaler. General-purpose image enhancement. From 2 credits.
- [Flux 3 Image](https://varg.ai/models/flux_3_image.md): `flux_3_image` — Black Forest Labs FLUX 3 Image. Newest image model — native 2K/4K, precise layout control, improved text rendering. Edits with up to 10 references. From 3 credits.
- [Flux 3 Image Klein](https://varg.ai/models/flux_3_image_klein.md): `flux_3_image_klein` — Black Forest Labs FLUX 3 Image Klein. Compact open-weights FLUX 3 Image (5.6B): native 4K, bounding-box layout, edits with up to 10 references. From 2 credits.
- [Flux Dev](https://varg.ai/models/flux_dev.md): `flux_dev` — Flux Dev. Good quality at lower cost. From 2 credits.
- [Flux Pro](https://varg.ai/models/flux_pro.md): `flux_pro` — Black Forest Labs Flux Pro 1.1. High-quality image generation. From 2 credits.
- [Flux Schnell](https://varg.ai/models/flux_schnell.md): `flux_schnell` — Flux Schnell. Fastest and cheapest Flux variant. From 2 credits.
- [Flux Schnell (PiAPI)](https://varg.ai/models/flux_schnell_piapi.md): `flux_schnell_piapi` — Flux Schnell via PiAPI provider. Alternative routing for Flux Schnell. From 2 credits.
- [gpt_image_1](https://varg.ai/models/gpt_image_1.md): `gpt_image_1` From 173 credits.
- [gpt_image_1_5](https://varg.ai/models/gpt_image_1_5.md): `gpt_image_1_5` From 105 credits.
- [gpt_image_1_mini](https://varg.ai/models/gpt_image_1_mini.md): `gpt_image_1_mini` From 105 credits.
- [gpt_image_2](https://varg.ai/models/gpt_image_2.md): `gpt_image_2` From 105 credits.
- [gpt_image_2_5](https://varg.ai/models/gpt_image_2_5.md): `gpt_image_2_5` From 173 credits.
- [gpt_image_2_5_flare](https://varg.ai/models/gpt_image_2_5_flare.md): `gpt_image_2_5_flare` From 105 credits.
- [gpt_image_2_5_sunburst](https://varg.ai/models/gpt_image_2_5_sunburst.md): `gpt_image_2_5_sunburst` From 105 credits.
- [Grok Imagine Image](https://varg.ai/models/grok_imagine_image.md): `grok_imagine_image` — xAI Grok image generation. Cheapest image model. From 2 credits.
- [Grok Imagine Image 2.0](https://varg.ai/models/grok_imagine_image_2.md): `grok_imagine_image_2` — Grok Imagine Image 2.0 by xAI. Top-ranked image generation and editing with strong text rendering, up to 2K. From 5 credits.
- [Grok Imagine Image Edit](https://varg.ai/models/grok_imagine_image%2Fedit.md): `grok_imagine_image/edit` — xAI Grok image editing. Cheapest image editing model. From 2 credits.
- [Ideogram 4.5](https://varg.ai/models/ideogram_v4_5.md): `ideogram_v4_5` — Posters, logos and designs with accurate text. Attach an image to edit it — up to 4 references, an optional mask, and a high-precision mode that keeps untouched pixels. From 2 credits.
- [Krea (latest)](https://varg.ai/models/krea.md): `krea` — The latest Krea image model — currently Krea 3. Pin krea_3 to stay on a version. From 0 credits.
- [Krea 2](https://varg.ai/models/krea_2.md): `krea_2` — Krea 2 Large. Krea's flagship image model: photoreal or stylised, steered by style-reference images, styles and moodboards. From 7 credits.
- [Krea 2 Medium](https://varg.ai/models/krea_2_medium.md): `krea_2_medium` — Krea 2 Medium. Krea's image model at half the price of Large, with the same style references, styles and moodboards. From 4 credits.
- [Krea 2 Medium Turbo](https://varg.ai/models/krea_2_medium_turbo.md): `krea_2_medium_turbo` — Krea 2 Medium Turbo. The fastest, cheapest Krea 2 tier with the same style references, styles and moodboards. From 2 credits.
- [Krea 3](https://varg.ai/models/krea_3.md): `krea_3` — Krea 3 (early access). Krea's newest image model — text-to-image and edits with up to 8 input images, fast / high / ultra thinking, 1K or 2K. From 0 credits.
- [Krea 2 Turbo Style](https://varg.ai/models/krea-2-turbo-style.md): `krea-2-turbo-style` From 2 credits.
- [mai_image_2_5](https://varg.ai/models/mai_image_2_5.md): `mai_image_2_5` From 21 credits.
- [Meta Muse Image](https://varg.ai/models/muse_image.md): `muse_image` — Muse Image by Meta. Top-10 image generation and editing for one cent an image; looks things up and runs code for accurate charts, QR codes and diagrams. From 2 credits.
- [Nano Banana 2](https://varg.ai/models/nano_banana_2.md): `nano_banana_2` — Google Gemini 2 Flash Image. Fast, cost-effective image generation. From 2 credits.
- [Nano Banana 2.1](https://varg.ai/models/nano_banana_2_1.md): `nano_banana_2_1` — Nano Banana 2.1 by Google. Top-ranked image generation and editing: sharper layouts and text, consistent subjects, 1K-4K and extreme aspect ratios. From 6 credits.
- [Nano Banana 2 Lite](https://varg.ai/models/nano_banana_2_lite.md): `nano_banana_2_lite` From 5 credits.
- [Nano Banana 2 Edit](https://varg.ai/models/nano_banana_2%2Fedit.md): `nano_banana_2/edit` — Google Gemini 2 Flash image editing. Reference-based generation. From 2 credits.
- [Nano Banana Pro](https://varg.ai/models/nano_banana_pro.md): `nano_banana_pro` — Google Gemini 3 Pro Image. Text-to-image generation. From 2 credits.
- [Nano Banana Pro Edit](https://varg.ai/models/nano_banana_pro%2Fedit.md): `nano_banana_pro/edit` — Google Gemini 3 Pro Image editing. Reference-based generation. From 2 credits.
- [Phota](https://varg.ai/models/phota.md): `phota` — Phota image generation. High-quality photorealistic output. From 2 credits.
- [Phota Edit](https://varg.ai/models/phota%2Fedit.md): `phota/edit` — Phota image editing. Edit existing images with text prompts. From 2 credits.
- [Phota Enhance](https://varg.ai/models/phota%2Fenhance.md): `phota/enhance` — Phota image enhancement. Improve quality and detail of existing images. From 2 credits.
- [Qwen Angles](https://varg.ai/models/qwen_angles.md): `qwen_angles` — Qwen multi-angle camera manipulation. Change viewing angle of a subject. From 2 credits.
- [Qwen Image 3](https://varg.ai/models/qwen_image_3.md): `qwen_image_3` — Qwen-Image 3 by Alibaba. Dense, accurate text in 12 languages, long prompts, comics and infographics up to 2048 px; edits combine up to three images. From 5 credits.
- [Recraft Clarity](https://varg.ai/models/recraft_clarity.md): `recraft_clarity` — Recraft Clarity upscaler. Enhance images with design-aware upscaling. From 2 credits.
- [Recraft v3](https://varg.ai/models/recraft_v3.md): `recraft_v3` — Recraft V3. Strong at icons, illustrations, and brand assets. From 2 credits.
- [Recraft V4.1 Flash](https://varg.ai/models/recraft_v4_1_flash.md): `recraft_v4_1_flash` — Recraft V4.1 Flash. Fast, very low-cost images from text — photography, illustration, mixed media — with color palette and background color control. From 2 credits.
- [Seedream 5.0 Flash](https://varg.ai/models/seedream_5_flash.md): `seedream_5_flash` — ByteDance Seedream 5.0 Flash. Fast, low-cost text-to-image and editing with up to 10 reference images, up to 2K. From 18 credits.
- [Seedream 5 Lite](https://varg.ai/models/seedream_5_lite.md): `seedream_5_lite` From 4 credits.
- [seedream_5_pro](https://varg.ai/models/seedream_5_pro.md): `seedream_5_pro` From 16 credits.
- [SeedDream v4.5 Edit](https://varg.ai/models/seedream_v4_5%2Fedit.md): `seedream_v4_5/edit` — ByteDance SeedDream. Image editing via text prompt. From 2 credits.
- [SeedVR Image Upscale](https://varg.ai/models/seedvr.md): `seedvr` — ByteDance SeedVR image upscaler. Good quality at moderate cost. From 2 credits.
- [Soul](https://varg.ai/models/soul.md): `soul` — Higgsfield Soul. 100+ character styles. Perfect for UGC creators. From 6 credits.
- [Topaz Image Upscale](https://varg.ai/models/topaz.md): `topaz` — Topaz-quality image upscaling. Best image enhancement quality. From 2 credits.

## Speech (tool: speech)

- [ElevenLabs Flash v2](https://varg.ai/models/eleven_flash_v2.md): `eleven_flash_v2` — Previous-gen flash model. Good baseline TTS. From 2 credits.
- [ElevenLabs Flash v2.5](https://varg.ai/models/eleven_flash_v2_5.md): `eleven_flash_v2_5` — Flash variant. Fastest TTS with solid quality. From 2 credits.
- [ElevenLabs Multilingual v2](https://varg.ai/models/eleven_multilingual_v2.md): `eleven_multilingual_v2` — Multilingual support. 29 languages with natural prosody. From 2 credits.
- [Eleven Turbo](https://varg.ai/models/eleven_turbo.md): `eleven_turbo` From 2 credits.
- [ElevenLabs Turbo v2](https://varg.ai/models/eleven_turbo_v2.md): `eleven_turbo_v2` — Previous-gen turbo model. Stable and reliable. From 2 credits.
- [ElevenLabs Turbo v2.5](https://varg.ai/models/eleven_turbo_v2_5.md): `eleven_turbo_v2_5` — Turbo speed. Low latency, good for real-time applications. From 2 credits.
- [ElevenLabs v3](https://varg.ai/models/eleven_v3.md): `eleven_v3` — Latest ElevenLabs model. Most expressive and natural. From 2 credits.
- [ElevenLabs v4](https://varg.ai/models/eleven_v4.md): `eleven_v4` — ElevenLabs v4. The most natural, expressive speech: top of the TTS arena, 90+ languages, audio tags and voice clones. From 2 credits.
- [ElevenLabs v4 Turbo](https://varg.ai/models/eleven_v4_turbo.md): `eleven_v4_turbo` — ElevenLabs v4 Turbo. Real-time (~100 ms) expressive speech at half the price of v4; #1 on the TTS arena. From 2 credits.
- [Gemini 3.8 Flash-Lite TTS](https://varg.ai/models/gemini_3_8_flash_lite_tts.md): `gemini_3_8_flash_lite_tts` — Gemini 3.8 Flash-Lite TTS by Google. Low-cost natural speech in 101 languages with 30 voices and two-speaker dialogue. From 2 credits.
- [Gemini 3.8 Flash TTS](https://varg.ai/models/gemini_3_8_flash_tts.md): `gemini_3_8_flash_tts` — Gemini 3.8 Flash TTS by Google. Natural speech in 130 languages with 30 voices, style instructions and two-speaker dialogue. From 2 credits.
- [Turbo](https://varg.ai/models/turbo.md): `turbo` From 2 credits.

## Music (tool: music)

- [Eleven Music](https://varg.ai/models/eleven_music.md): `eleven_music` From 2 credits.
- [ElevenLabs Music v2.5](https://varg.ai/models/eleven_music_v2_5.md): `eleven_music_v2_5` — ElevenLabs Music v2.5. Studio-grade songs with vocals or instrumentals, 3 seconds to 10 minutes. From 2 credits.
- [Lyria 3.5](https://varg.ai/models/lyria_3_5.md): `lyria_3_5` — Lyria 3.5 by Google. Full songs up to about 3 minutes with vocals and lyrics, or instrumentals; set length and structure in the prompt. From 11 credits.
- [Mureka 9.5](https://varg.ai/models/mureka_9_5.md): `mureka_9_5` — Mureka 9.5 instrumental generation: a text description of up to 1,024 characters in, an instrumental track out. No vocals. Older Mureka 9, 8 and 7.6 are selectable through provider_options. From 6 credits.
- [ElevenLabs Music](https://varg.ai/models/music_v1.md): `music_v1` — AI music generation. Prompt-driven background tracks. From 2 credits.

## Transcription (tool: transcription)

- [Groq Whisper](https://varg.ai/models/groq_whisper.md): `groq_whisper` — Groq-hosted Whisper. Faster inference. From 53 credits.
- [Groq Whisper Large v3](https://varg.ai/models/groq_whisper_large_v3.md): `groq_whisper_large_v3` — Groq-hosted Whisper Large v3. Fast inference with high accuracy. From 105 credits.
- [Groq Whisper Turbo](https://varg.ai/models/groq_whisper_large_v3_turbo.md): `groq_whisper_large_v3_turbo` — Fastest and cheapest transcription. Groq-accelerated Whisper. From 53 credits.
- [Whisper](https://varg.ai/models/whisper.md): `whisper` — OpenAI Whisper via fal. Accurate speech-to-text. From 6 credits.
- [Whisper Large v3](https://varg.ai/models/whisper_large_v3.md): `whisper_large_v3` — OpenAI Whisper Large v3 via fal. Higher accuracy for complex audio. From 6 credits.

## Media editing (tool: ffmpeg)

- [Ffmpeg](https://varg.ai/models/ffmpeg.md): `ffmpeg` From 6 credits.
- [Resize](https://varg.ai/models/resize.md): `resize` From 6 credits.
- [Slice](https://varg.ai/models/slice.md): `slice` From 6 credits.
- [Trim](https://varg.ai/models/trim.md): `trim` From 6 credits.

## Video rendering (tool: render)

- [Render](https://varg.ai/models/render.md): `render` From 6 credits.

## Video_analyze (tool: video_analyze)

- [video-analyze](https://varg.ai/models/video-analyze.md): `video-analyze` From 6 credits.

## Optional

- [Site overview](https://varg.ai/llms.txt)
- [Docs index](https://docs.varg.ai/llms.txt)
- [OpenAPI](https://docs.varg.ai/openapi.yaml)