AI models

Pick a model. Call it. Switch any time.

This is every model you can call through OneInfer: 431 text, image, video, audio and speech models from 59 model makers, with prices loaded live from the catalog per token, image, second of video or request. One API and one key reach them all, so switching between models of the same type means changing one parameter.

To choose, search by name or tag, filter by provider, modality, context length, capability and price, then sort by price or context. Open a model for its specs and full rates. The filters stay in the address bar, so you can share the list.

Featured models

1 of 6

All models

431 models

Decision-1.1 logo

Microsoft: Microsoft-Decision-1.1

Microsoft

New

Microsoft-Decision-1.1 is Microsoft's decision model from November 2026, the successor to Microsoft-Decision-1. It reads a state (text, JSON or PNG and JPEG images) and typed questions (yes/no, choice or score) and returns a calibrated probability for every option, with no text generated. It has a 32K-token context window and is billed by input tokens only.

Input cost$0.042 / 1M tokens
Output cost$0.000 / 1M tokens
Context length32K
MAI-Image-2.7-Flash logo

Microsoft: MAI Image 2.7 Flash

Microsoft

New

MAI-Image-2.7 Flash is the faster, lower-cost member of Microsoft AI's MAI-Image-2.7 family from November 2026, for high-volume image generation and editing. It takes up to five reference images, can make transparent backgrounds, and is billed by the text and image tokens each request uses.

Text input tokens$1.75 / 1M tokens
Image input tokens$2.50 / 1M tokens
Image output tokens$10.00 / 1M tokens
Context lengthN/A
MAI-Image-2.7 logo

Microsoft: MAI Image 2.7

Microsoft

New

MAI-Image-2.7 is Microsoft AI's image generation and editing model from November 2026, the precision tier of the MAI-Image-2.7 family. It takes up to five reference images, can make transparent backgrounds, and is billed by the text and image tokens each request uses.

Text input tokens$5.00 / 1M tokens
Image input tokens$8.00 / 1M tokens
Image output tokens$34.00 / 1M tokens
Context lengthN/A
gpt-6.2-sol logo

OpenAI: GPT-6.2 Sol

OpenAI

New

GPT-6.2 Sol is OpenAI's everyday model for coding and agents at scale, released 17 November 2026 at GPT-6.1 Sol's price. It reads text and images, reasons with adjustable effort, calls tools such as web search and computer use through the Responses API and has a 1.05M-token context window with up to 128K output tokens. Prompts over 272K tokens are billed at OpenAI's long-context rates.

Input cost$2.000 / 1M tokens
Output cost$10.000 / 1M tokens
Context length1.05M
gpt-6.2-astra logo

OpenAI: GPT-6.2 Astra

OpenAI

New

GPT-6.2 Astra is OpenAI's most capable model for the hardest projects, released 17 November 2026: it builds and revises software, uses a computer and works with documents, with input and output 20% below GPT-6 Astra's prices. It reads text and images, reasons with adjustable effort, calls tools through the Responses API and has a 1.05M-token context window with up to 128K output tokens. Prompts over 272K tokens are billed at OpenAI's long-context rates.

Input cost$8.000 / 1M tokens
Output cost$40.000 / 1M tokens
Context length1.05M
dreamina-seedance-2-5-flash-261115 logo

ByteDance: Seedance 2.5 Flash

ByteDance

New

Seedance 2.5 Flash is ByteDance's faster, lower-cost Seedance 2.5 video model, released in November 2026. It makes 480p or 720p clips with sound from text, a first (and last) frame or reference images, video and audio, as Seedance 2.5 does. Billed by the video tokens each clip uses: about $0.07 a second at 480p and $0.15 at 720p (16:9), less per token when the request includes a video.

Video generationfrom $0.0673 / second
Pricing options2 tiers
ltx-alpha-gen logo

Lightricks: LTX Alpha Gen

Lightricks

New

LTX Alpha Gen separates a video's foreground from its background. It returns a grayscale matte MP4 and a ProRes 4444 cutout with the matte as its alpha channel, frame for frame, keeping soft edges like hair. Send the video in video_url: up to 4K and 3,072 frames (about 2 minutes at 24 FPS). Billed per second of the input video, by its short side (1080p and below, 1440p, or 4K).

Video generationfrom $0.1000 / second
Pricing options3 tiers
ltx-upscale logo

Lightricks: LTX Upscale

Lightricks

New

LTX Upscale raises a video's resolution to 1080p, 4K or 8K (set resolution), keeping its aspect ratio, frame rate and sound. Send the video in video_url: H.264 or HEVC, at least 480 pixels on the short side, up to 2401 frames for 1080p, 601 for 4K or 241 for 8K (about 100, 25 or 10 seconds at 24 FPS). Billed per second of the input video, by the output resolution.

Video generationfrom $0.0800 / second
Pricing options3 tiers
dola-seedream-5-5-261028 logo

ByteDance: Seedream 5.5

ByteDance

New

Seedream 5.5 is ByteDance's image generation and editing model, released in November 2026. It makes one image per request at 1K, 1.5K or 2K from a prompt and up to 10 reference images, with steadier multi-turn edits than Seedream 5.0. Billed per image by its size, plus each reference image.

Price per imagefrom $0.02 / image
Price per input image$0.0080 / image
Context lengthN/A
scribe_v3 logo

ElevenLabs: Scribe v3

ElevenLabs

New

Scribe v3 is ElevenLabs' speech-to-text model, released 10 November 2026: the successor to Scribe v2, with 27% fewer word errors across 102 languages (FLEURS). It takes Scribe v2's options, plus secondary_language_codes for audio in several languages and numbers_format to write numbers as digits or words. Billed per hour of audio, as Scribe v2.

Speech transcription$0.220 / hour
Context lengthN/A
gemini-decision-1 logo

Google: Gemini Decision 1

Google

New

Gemini Decision 1 is Google's decision model, released 11 November 2026. Given a state (text or JSON) and typed questions (yes/no, choice or score), it returns a calibrated probability and a confidence for every question in one forward pass, with no text generated. It reads up to 256K tokens, shared by the state and the questions.

Input cost$0.100 / 1M tokens
Output cost$0.000 / 1M tokens
Context length256K
gemini-4-hydrogen logo

Google: Gemini 4 Hydrogen

Google

New

Gemini 4 Hydrogen is Google's fast Gemini 4 model for everyday coding and agentic work, released 12 November 2026; Google reports it scores above Gemini 4 Argon. It reads text, images, audio, video and PDFs, thinks at low, medium, high (the default) or xhigh level, calls functions and has a 1M-token context window, priced flat across all of it.

Input cost$1.500 / 1M tokens
Output cost$7.500 / 1M tokens
Context length1M

About the model catalog

Compare the models above by type, provider and current price, then call any of them through one API, with no infrastructure to manage. Within a model type, switching models means changing the model parameter in your request.

Text and language models

The catalog includes large language models such as GPT-6.1 Terra, Claude Opus 5.5, Claude Haiku 5.5, Gemini 4 Argon, Kimi K3.1 Pro, Qwen3.8 Max, DeepSeek V4 Pro, Mistral Large 4 and GLM-5.3. Text models are called through the same chat completions endpoint and billed per token. For steady production traffic, you can also run a model on GPUs reserved for you with a model deployment.

Vision and image models

Multimodal models such as Gemini 4 Argon, GPT-6.1 Terra, Kimi K3.1 Pro and Qwen3.8 Flash accept images alongside text for image understanding, document analysis and visual question answering. Image generation models such as GPT Image 2.5, FLUX 3 Image, Seedream 5.0 Pro and Imagen 3 are called through the image generation endpoint.

Audio and speech models

Speech-to-text models such as ElevenLabs Scribe v2 and Microsoft MAI Transcribe 2 handle transcription. Text-to-speech models such as ElevenLabs Eleven v4, MiniMax Speech 2.8 HD and Gemini 3.8 Flash TTS turn text into speech in a choice of voices.

Video generation models

Text-to-video and image-to-video models such as Kling 4.0, Seedance 2.5, Wan 3.0, Veo 3.1 and Gemini Omni 1.1 Flash are called through the video generation endpoint. Each model card shows how that model is priced, for example per second or per video.

How pricing works

Each model card in the catalog above shows that model's current price: per token, per image, per second of video or per request, depending on the model. You pay as you go from prepaid API credits.

Prices are loaded from the live catalog, so they change when providers change their rates. For high-volume workloads, contact us about model deployment pricing.