# Qwen3.8 2.4T A95B API prices

Qwen3.8 2.4T A95B (qwen) — 1.0M context, 1.0M max output.
21 metered per-token offers from 18 provider(s), in USD per 1M tokens.
Prices as of 2026-10-04.

| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context |
| --- | --- | --- | --- | --- |
| IteraCompute | $1.95 | $5.95 | $0.2 | 970k |
| AIHubMix | $2 | $6 | $0.5 | 262k |
| Deep Infra | $2 | $6 | $0.2 | 262k |
| Eden AI | $2 | $6 | $0.25 | 1.0M |
| Fireworks AI | $2 | $6 | $0.25 | 262k |
| Charm Hyper | $2 | $6 | $0.25 | 1.0M |
| Kilo Gateway | $2 | $6 | $0.25 | 1.0M |
| DevPass (LLM Gateway) | $2 | $6 | $0.25 | 1.0M |
| LLM Gateway | $2 | $6 | $0.2 | 262k |
| LLM Gateway | $2 | $6 | $0.25 | 1.0M |
| LLM Gateway | $2 | $6 | $0.25 | 1.0M |
| OpenRouter | $2 | $6 | $0.25 | 1.0M |
| Requesty | $2 | $6 | $0.2 | 262k |
| SiliconFlow | $2 | $6 | $0.25 | 1.0M |
| Vercel AI Gateway | $2 | $6 | $0.25 | 262k |
| Cortecs | $2.5 | $6 | $0.625 | 262k |
| Opper | $2.5 | $6 | $0.63 | 262k |
| Requesty (eu) | $2.5 | $6 | $0.63 | 1.0M |
| TensorX | $2.5 | $6 | $0.63 | 262k |
| Hugging Face | $2.5 | $6.25 | — | 262k |
| Merge Gateway | $2.5 | $6.25 | $0.5 | 262k |

## Summary

- Cheapest input: $1.95 per 1M tokens (IteraCompute)
- Cheapest output: $5.95 per 1M tokens (IteraCompute)
- First-party: not listed separately
- Free offers: none
- Subscription plans (not per-token): none
- Released: 2026-08-12
- Inputs: image, text

## FAQ

### What is the cheapest Qwen3.8 2.4T A95B API?

As of Oct 4, 2026, IteraCompute has the lowest Qwen3.8 2.4T A95B output price at $5.95 per 1M tokens, and IteraCompute has the lowest input price at $1.95 per 1M tokens. Prices are metered per-token rates in USD; free tiers and subscription plans are excluded.

### How many providers offer Qwen3.8 2.4T A95B?

18 providers list Qwen3.8 2.4T A95B on Sovyron; 21 of them sell it at a metered per-token price.

### Is Qwen3.8 2.4T A95B free?

No provider in the Sovyron catalog lists a free tier for Qwen3.8 2.4T A95B.

### What is the context window of Qwen3.8 2.4T A95B?

Qwen3.8 2.4T A95B supports a 1.0M-token context window and up to 1.0M output tokens.

HTML page: https://sovyron.com/models/qwen3-8-2-4t-a95b/
Full dataset: https://sovyron.com/data/catalog.json