# GPT-6 Sol and Luna now available on AI Gateway

Source: [Vercel](https://vercel.com/changelog/gpt-6-sol-and-luna-now-available-on-ai-gateway)  
Feed7 permalink: https://feed7.dev/p/gpt-6-sol-and-luna-now-available-on-ai-gateway-11bfugh  
Published: 2026-09-22T00:00:00.000Z  
Trust: Official Source (official_source)

## Why Included

Vercel AI Gateway now exposes GPT-6 Sol for demanding coding runs and Luna for high-volume work, with unified usage tracking, retries, failover, and routing.

## Source Summary

Vercel AI Gateway now offers **openai/gpt-6-sol** for complex, sustained coding work and **openai/gpt-6-luna** for lower-cost, high-volume agent workflows. Both are available through the AI SDK, compatible Chat Completions and Responses APIs, and connected coding agents.

## Practical Implication

Use Sol where longer tasks and iteration quality matter; test Luna for throughput-sensitive work. Gateway routing, usage tracking, retries, and failover make it practical to evaluate both behind one integration.

## Agent-Ready Context

Vercel AI Gateway now offers **openai/gpt-6-sol** for complex, sustained coding work and **openai/gpt-6-luna** for lower-cost, high-volume agent workflows. Both are available through the AI SDK, compatible Chat Completions and Responses APIs, and connected coding agents.

Use Sol where longer tasks and iteration quality matter; test Luna for throughput-sensitive work. Gateway routing, usage tracking, retries, and failover make it practical to evaluate both behind one integration.

Vercel says both cost less than GPT-6 Astra and improve factual reliability over GPT-5.6 counterparts, but the material includes no prices, benchmark results, or quantified reliability gains.

## Connected Context

Feed7 judgment across 856 accumulated Signals:

This advances the earlier GPT tiering pattern from GPT-5.6 to GPT-6, preserving a practical split between sustained coding quality and high-volume economy behind the same gateway integration. It expands routing options but does not justify replacing existing defaults: the lower-cost and reliability claims lack prices, benchmarks, and quantified gains, so matched workload evaluation remains necessary.

- [AI Gateway adds unified fast mode support](https://feed7.dev/p/ai-gateway-adds-unified-fast-mode-support-144dq26) — Unified fast mode supplies a separate latency control that can complement model-tier routing, so evaluations should distinguish gains from choosing Luna or Sol from gains produced by serving mode.
- [GLM 5.3 FlashX now available on AI Gateway](https://feed7.dev/p/glm-5-3-flashx-now-available-on-ai-gateway-1yir9b2) — GLM 5.3 FlashX offers a concrete throughput claim, whereas GPT-6 Luna is positioned for volume without quantified speed; this makes end-to-end latency and quality testing more useful than labels alone.

## Context Map

- Layer: infra
- Domains: coding
- Topics: gateways, model-selection, coding-agents

## Uncertainty

- Vercel says both cost less than GPT-6 Astra and improve factual reliability over GPT-5.6 counterparts, but the material includes no prices, benchmark results, or quantified reliability gains.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.