InclusionAI's Ling 3.1 Flash with 1M-token context window arrives free on Vercel AI Gateway

What launched

Vercel has added Ling 3.1 Flash 1019, a language model from InclusionAI, to its AI Gateway product, and is making it free to use until November 3, 2026. According to the Vercel blog, requests made during the promotional window will still appear in the spend dashboard with a full request trace — they simply cost nothing. Once the window closes, requests will begin charging at the model's published rate.

A long-context model for coding and agents

Vercel describes the updated model as built for three related workloads: coding, multi-step reasoning, and tool-using agents. Its defining specification is a 1M-token context window, which the company says suits work with lengthy documents, codebases, and long-running task histories — the accumulated state that builds up when an agent works through a multi-step job rather than answering a single prompt.

On the gateway, the model is addressed as inclusionai/ling-3.1-flash-1019.

How to try it

Developers who want to route the model through coding agents such as Claude Code, Codex, or Cursor can install the latest Vercel CLI and run its setup command, npx vercel ai-gateway setup, then select inclusionai/ling-3.1-flash-1019 inside the agent. Vercel also offers the model in its model playground for quick evaluation without wiring up a client, and points to its coding agents guide for further details.

What AI Gateway provides

AI Gateway is Vercel's unified API for calling language models. Beyond request routing, it handles usage and cost tracking, retries, failover, and performance optimizations that Vercel says deliver higher uptime than calling providers directly. The service also includes built-in custom reporting, Zero Data Retention support, budgets for API keys, and routing rules.

On pricing, Vercel states that AI Gateway reflects provider pricing with no markup and charges no platform fee on inference — including on Bring Your Own Key (BYOK) requests, where users supply credentials for a provider they already pay.

Why it matters

A free window on a 1M-token context model is a low-cost way for teams to stress-test the workloads where long context actually earns its keep: agentic coding sessions that accumulate tool calls and file reads, analysis over lengthy documents, and reasoning chains that outgrow smaller windows. Because free requests still produce full traces in the spend dashboard, the promotion doubles as a low-friction evaluation of AI Gateway itself, not just of the model.

The launch also illustrates how this layer of the AI stack is competing. Vercel is folding more models into a single API with transparent pass-through pricing, and using time-limited free access as an acquisition lever. For developers, the practical caveat is the deadline: anything that proves useful before the cutoff should be budgeted at the published rate afterward, since free access ends on November 3, 2026.