Updates

New models, improvements and fixes. Newest first.

  1. New

    Now available: Studio — generate videos and voiceovers

    A new Studio page turns a text prompt into a video or a voiceover. Nine video models and three voice models, paid from your prepaid balance.

    Studio is a new page in your dashboard for generating videos and voiceovers from a text prompt. Pick a model, describe what you want, and the result appears in your gallery when it is ready — usually a few minutes for video, under a minute for voice.

    Video models

    Veo 3.1 — $0.03 per 8-second clip

    Veo 3.1 1080p — $0.075 per 8-second clip

    Omni 1.1 Flash — $0.02 per second

    Grok Imagine 720p — $0.035 per second

    Seedance 2.0 — $0.165 per second · 1080p $0.305 per second

    Seedance 2.5 — 480p $0.15 · 720p $0.265 · 1080p $0.61 per second

    Start a clip from your own image, choose landscape (16:9) or portrait (9:16), and set the length on models that support it.

    Voice models — $0.045 per 1,000 characters

    Eleven v4, Eleven v3 and Eleven Multilingual v2, with a choice of voices and languages. Every voiceover is delivered as an MP3.

    Charges come from your prepaid balance, which never expires, and a render that fails is never charged.

    Open Studio

    Building your own app? The same models are available through the API: send POST https://omnirouter.li/v1/jobs with your API key, then poll GET https://omnirouter.li/v1/jobs/{id} until the render is ready.

  2. New

    Now available: GPT 6.1 Sol and Sonnet 5.5

    Two new models are live in the catalog, both supporting streaming and image input.

    Two models are now available in the catalog.

    GPT 6.1 Sol — $0.27 per 1M input · $1.20 per 1M output

    Sonnet 5.5 — $0.50 per 1M input · $1.50 per 1M output

    Both support streaming and accept image input. Charges are per request, from your prepaid balance — which never expires.

    Each model is routed through multiple upstream providers with automatic failover. You are charged the listed price for the model regardless of which provider serves the request.

    Point your existing [OI]-compatible client at https://omnirouter.li/v1 and use the model identifier from the catalog.

  3. New

    Now available: Opus 5.5, GPT 6 Sol, and GPT 6 Luna

    Three frontier-class models join the catalog, priced from $0.010 per 1M input tokens.

    Three models are now available in the catalog.

    Opus 5.5 — $0.70 per 1M input · $3.20 per 1M output GPT 6 Sol — $0.18 per 1M input · $0.90 per 1M output GPT 6 Luna — $0.010 per 1M input · $0.050 per 1M output

    All three support streaming. Charges are per request, from your prepaid balance — which never expires.

    Each model is routed through multiple upstream providers with automatic failover, so one slow provider does not take the model offline. You are charged the listed price for the model regardless of which provider serves the request.

    The catalog shows each model's listed price alongside the vendor's published comparison rate. Point your existing [OI]-compatible client at https://omnirouter.li/v1 and use the model identifier from the catalog.