AI Agent StoreParticle AI · Provider and hosting offer verified
Particle-1
Try a coding model trained to finish software tasks with fewer generated tokens. Compare finished work, retries and the whole bill on your own repositories.
Released 2026-10-28 · Catalog added 2026-10-28 UTC · Checked 2026-11-03. Catalog addition and provider launch are separate dates.
We still need comparable independent tests. You can compare the confirmed features and limits below, but we cannot yet give this version an Agent Fit or Practical Value rating. Test it on your own tasks before switching; results for another version or thinking setting may not apply.
What you can use today
Token prices measure the small pieces of text the model reads or produces. Capacity is the amount the service accepts in a request; check the limits below for any fallback to another model. A dash means the rate is unverified.
Exact catalog ID: particle-ai/particle-1. Adapted from DeepSeek V4.1 Flash for coding and data-engineering tasks. It accepts text and images and can call tools. Publisher efficiency tests do not establish an independent quality score or guaranteed savings.
The native request alias is particle-1; responses identify particle-ai/particle-1. Native thinking defaults to high: none disables it; minimal/low map to low, medium/high/xhigh to high, and max/ultra to max. Gateway lists none/low/high/max separately. A separate US hostname restricts inference GPUs to the United States. It does not restrict file storage, billing records or other services; retries must stay on the US hostname.
Native and routed documentation list 1,048,576 tokens shared by input and output, with a 131,072 output ceiling. Native requests default to 32,000 output tokens when no limit is supplied. Thinking uses the same output budget.
Native standard rates are $0.25 input, $0.03 cached input and $1 output per million tokens. Off-peak rates are half, from 20:30 to 04:00 America/New_York. Effective Flex halves token prices again; requesting Flex alone does not guarantee entitlement. Web searches cost extra. Router and Gateway quotes apply to their own offers. Allow for retries and any tool fees. Models that think before answering may also charge for that work.
Provider release and specifications ↗
Inspect the exact catalog model ↗
More setup and licensing details ↗
More setup and licensing details ↗
More setup and licensing details ↗
More setup and licensing details ↗
Test it before switching
Give each model the same examples, tools and definition of success. Record correct outcomes, response time, retries and total cost. A cheaper token price does not guarantee a cheaper completed task.
Compare caching and long-context costs ↗