OpenRouter has released a free AI model called Crouton Alpha, which supports a 1 million token context window and accepts text, images, audio, and video, while its developer has not been publicly identified.
OpenRouter · Free model release
Crouton Alpha: a free, 1M-token multimodal model with no named developer
OpenRouter is offering Crouton Alpha at no cost, aimed at high-volume agent workloads. Its builder has not been publicly identified, leaving key questions about performance, safety, and terms unanswered.
1M
token context window for a single session
Free
no-cost access, duration not stated
4
input types: text, images, audio, video
None
developer named in the announcement
What 1M tokens means in blocks
Illustrative scale: each block is 100,000 tokens, so the full context window is ten blocks.
1,000,000 tokens = one session's worth of material, such as long documents, logs, or media combined in a single pass.
The figure describes how much the model can consider at once. It does not measure accuracy or reliability.
What is confirmed
Free to use on OpenRouter
1M-token context window
Accepts text, images, audio, and video
Provider says it does not train on user data
Positioned for repetitive, agent-driven work
What is still open
Who built the model and how it was trained
Architecture, training data, and benchmarks
Safety testing results
How long it stays free
Rate limits and usage restrictions
Why developers are paying attention
Step 1
Long inputs in one pass
Documents, logs, and media
→
Step 2
Agent workflows
High-volume, repetitive tasks
→
Step 3
Test at no spend
Free access lowers trial cost
→
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...