Models / LLMs with video input

LLMs with video input

Every generally-available model that can watch — take a video clip as input to summarise it, answer questions about it, or extract information from the footage — ranked by blended price per 1M tokens. Sortable and filterable.

The short answer

We track 13 generally-available models that can watch — models that accept video input — from 4 providers. The most affordable by published token price is Amazon Nova Lite (Amazon) at $0.105 per 1M tokens blended. The largest context window among them is Gemini 3.1 Flash-Lite at 1.0M tokens.

Last verified: 9 Jul 2026 · sourced from official provider documentation

Most affordable video-capable modelAmazon Nova Lite
$0.105/1M blended · $0.06 in / $0.24 out · 300K context · Amazon

These models understand video you send them — they are not video generators. Text-to-video models (Sora, Veo, Grok Imagine) are a separate class and are excluded so the price comparison stays like-for-like. Video input is usually billed by sampling frames into image tokens; the figures shown are the model’s standard token rates, so confirm the per-second/per-frame cost on the provider’s pricing page before budgeting.

The list

13 models that can watch

Generally-available models only, ranked by blended $/1M tokens. Click a header to re-sort, or filter by provider.

Model Status Context Input $/M Output $/M Blended $/M Cutoff
Amazon Nova Lite
Amazon · text, image, video
GA300K$0.06$0.24$0.1052024-10
Gemini 3.1 Flash-Lite
Google · text, image, video, audio, pdf
GA1.0M$0.25$1.50$0.5632025-01
Qwen3.6-Flash
Alibaba · text, image, video
GA1M$0.25$1.50$0.563
Qwen3.7-Plus
Alibaba · text, image, video
GA1M$0.4$1.60$0.7
Qwen3.7-Plus (snapshot 2026-05-26)
Alibaba · text, image, video
GA1M$0.4$1.60$0.7
Amazon Nova 2 Lite
Amazon · text, image, video
GA1M$0.3$2.50$0.852025-10
Qwen3.6-Plus
Alibaba · text, image, video
GA1M$0.5$3$1.13
Kimi K2.5
Moonshot · text, image, video
GA262K$0.6$3$1.20
Amazon Nova Pro
Amazon · text, image, video
GA300K$0.8$3.20$1.402024-10
Kimi K2.6
Moonshot · text, image, video
GA262K$0.95$4$1.71
Gemini 3.5 Flash
Google · text, image, video, audio, pdf
GA1.0M$1.50$9$3.382025-01
Qwen3.7-Max
Alibaba · text, image, video
GA1M$2.50$7.50$3.75
Qwen3.5-Omni-Plus
Alibaba · text, image, audio, video
GA

Blended = 0.75 × input + 0.25 × output $/M tokens (a fair single-number cost proxy). Click any header to sort.

FAQ

LLMs with video input

Which LLMs accept video input?

We track 13 generally-available models that accept video input, across 4 providers — from Amazon Nova Lite (Amazon) at the low end to the flagships from Anthropic, OpenAI, Google and others. The full, filterable list is on this page, ranked by blended price. Prices come from official provider documentation and are refreshed daily.

What is the cheapest video model?

Amazon Nova Lite (Amazon) is the cheapest generally-available model we track that can watch, at $0.06 per 1M input tokens and $0.24 per 1M output tokens — a blended $0.105 per 1M. Cheapest on paper is not always best: check the quality, latency and rate limits before committing.

Do these models generate video?

No. This page lists models that accept video as input — video understanding, such as summarising a clip, answering questions about footage or reading text from a frame. Text-to-video generators (Sora, Veo, Grok Imagine) are a separate class and are excluded here so the comparison stays like-for-like on token price.

Is the cheapest video-capable model good enough for production?

Not always. Price is only one axis — quality, latency, rate limits and reliability matter just as much, and the cheapest model on paper can cost more once you account for retries or larger prompts. Use this ranking to build a shortlist, then test the top few on your own workload before committing.