# Qwen3.8 27B API prices

Qwen3.8 27B (qwen) — 1.1M context, 262k max output.
30 metered per-token offers from 31 provider(s), in USD per 1M tokens.

| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context |
| --- | --- | --- | --- | --- |
| Cortecs | $0.1 | $0.4 | $0.04 | 262k |
| RunInfra | $0.1 | $0.4 | $0.01 | 262k |
| EmpirioLabs AI | $0.17 | $0.5 | $0.08 | 262k |
| NanoGPT | $0.15 | $0.7 | $0.04 | 262k |
| CrofAI | $0.2 | $1.5 | $0.03 | 262k |
| DevPass (LLM Gateway) | $0.2 | $2 | $0.05 | 33k |
| LLM Tech | $0.25 | $2.09 | $0.04 | 262k |
| AKI.IO | $0.3 | $2.2 | $0.1 | 262k |
| Ofox | $0.5 | $1.71 | $0.043 | 1.1M |
| Kosmik Compute | $0.35 | $2.2 | $0.09 | 262k |
| OrcaRouter | $0.33 | $2.4 | — | 262k |
| IteraCompute | $0.3 | $2.5 | — | 328k |
| Kilo Gateway | $0.425 | $2.55 | $0.085 | 1.0M |
| Regolo AI | $0.58 | $2.42 | — | 120k |
| ai& | $0.4 | $3 | $0.2 | 262k |
| Deep Infra | $0.4 | $3 | $0.04 | 262k |
| Hugging Face | $0.4 | $3 | — | 262k |
| Weights & Biases | $0.4 | $3 | $0.15 | 262k |
| LLM Gateway | $0.42 | $3 | $0.085 | 1.0M |
| OpenRouter | $0.42 | $3 | $0.085 | 1.0M |
| Cerebras | $0.99 | $1.49 | — | 66k |
| Eden AI | $0.5 | $3 | $0.1 | 1.0M |
| Charm Hyper | $0.5 | $3 | $0.1 | 1.0M |
| Vercel AI Gateway | $0.5 | $3 | $0.1 | 1.0M |
| Cloudflare Workers AI | $0.45 | $3.2 | $0.05 | 262k |
| Neuralwatt | $0.45 | $3.2 | $0.25 | 262k |
| Ofox | $0.45 | $3.2 | $0.05 | 1.1M |
| Berget.AI | $0.46 | $3.48 | — | 262k |
| evroc | $0.87 | $3.5 | — | 262k |
| Groq | $0.8 | $4 | — | 131k |

## Summary

- Cheapest input: $0.1 per 1M tokens (Cortecs)
- Cheapest output: $0.4 per 1M tokens (Cortecs)
- First-party: not listed separately
- Free offers: Hetzner, OVHcloud AI Endpoints
- Subscription plans (not per-token): none
- Released: 2026-08-14
- Inputs: image, pdf, text, video

HTML page: https://llmprice.gitlab.io/models/qwen3-8-27b/
Full dataset: https://llmprice.gitlab.io/data/catalog.json