Particle-1 coding model now available on Vercel's AI Gateway
Vercel has added Particle-1, a model from Particle AI, to its AI Gateway. According to the changelog, the model is post-trained on DeepSeek V4.1 Flash and aims to complete software engineering tasks using fewer tokens.
- Particle-1 from Particle AI is now available on Vercel's AI Gateway.
- The model is post-trained on DeepSeek V4.1 Flash and targets software engineering tasks with fewer tokens.
- It has a 1M-token context window.
- The model name is particle/particle-1, and it works with the AI SDK, Chat Completions and coding agents via the Vercel CLI.
- Vercel says AI Gateway reflects provider pricing with no markup and no platform fee on inference.
A coding model built to use fewer tokens
Particle-1 is post-trained on DeepSeek V4.1 Flash, with a focus on software engineering work. Particle describes the gains as shorter reasoning, fewer rounds of trial and repair, and more focused exploration.
Vercel lists several suggested uses: debugging, planning code changes, and breaking down multi-step engineering problems. The model has a 1M-token context window, which Vercel says is meant for working with code and extended task histories.
How to call it
Developers need to set AI_GATEWAY_API_KEY in their environment and use particle/particle-1 as the model name. The changelog shows an example using the AI SDK's streamText function, which sends a prompt asking for three fixes for a flaky API retry loop and streams the response to the console. Chat Completions is also listed as a supported option.
For coding agents such as Claude Code, Codex and Cursor, Vercel says users should install the latest version of the Vercel CLI and run npx vercel ai-gateway setup. After that, they select particle/particle-1 in the agent. Vercel points to its coding agents guide for details and offers a model playground for trying Particle-1.
What AI Gateway offers
AI Gateway is described by Vercel as a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. Listed features include:
- Built-in custom reporting
- Zero Data Retention support
- Budgets for API keys
- Routing rules
On pricing, Vercel states that AI Gateway reflects provider pricing with no markup. It also says there is no platform fee on inference, including on Bring Your Own Key (BYOK) requests. The changelog does not give Particle-1 prices or benchmark results.
What it means for developers
Teams already routing model calls through AI Gateway can now try Particle-1 by changing the model name, without a separate integration. The token-efficiency focus is Particle's own claim, and the source offers no figures to measure it. Developers interested in the model can test it in the playground or in their coding agent and compare results on their own workloads.
Vercel: Particle-1 is now available on AI Gateway
