Alibaba (Qwen) › Qwen3.8 2.4T A95B › Pricing
Qwen3.8 2.4T A95B pricing
API list price by Alibaba (Qwen) · checked Oct 26, 2026
$2 / $6per 1M tokens in / out
$0.25cache read · −87%
$2.16per task at thinking?
#159on OpenRouter by tokens · 23B / week
Price sheetper 1M tokens
| Input | $2 | list |
| Output | $6 | |
| Cache read | $0.25 | −87% vs input |
| Cache write | $2.50 | +25% vs input |
| Context · max output | 262K · 131K | tokens |
Where to buy6 providers · per 1M tokens
| Provider | Input | Output | Context |
|---|---|---|---|
| Featherless | $1 | $4 | — |
| Alibaba | $2 | $6 | 1M |
| DeepInfra | $2 | $6 | — |
| Novita | $2 | $6 | — |
| SiliconFlow | $2 | $6 | 1M |
| Together | $2 | $6 | 1M |
In Alibaba (Qwen)'s line-upper 1M in / out
| Qwen3.5 397B A17B | $0.60 / $3.60 | −55% |
| Qwen3 Next 80B A3B Thinking | $0.50 / $6 | −37% |
| Qwen 3.5 Omni Plus 2026-03-15 | $0.96 / $5.50 | −30% |
| Qwen3.8 2.4T A95B | $2 / $6 | this model |
| Qwen3.8 Max (0902) | $2 / $6 | 0% |
| Qwen 2 VL 72B Instruct | $2.29 / $6.88 | +15% |
| Qwen MT Plus | $2.46 / $7.37 | +23% |
Cheaper for the same workvs Qwen 2.4T A95B
| Claude Opus 5.5 high | -15% | score +34% | $1.82 |
| Gemini 4 Argon high | -25% | score +32% | $1.62 |
| Claude Sonnet 5.5 xhigh | -7% | score +30% | $2.01 |
| GPT-6.1 Sol max | -66% | score +30% | $0.72 |
| GPT-6 Astra high | -20% | score +27% | $1.73 |
Pricing questions
How much does Qwen3.8 2.4T A95B cost?›
$2 per 1M input tokens and $6 per 1M output tokens. Cached input is $0.25. One Intelligence Index task costs about $2.16 at the default effort.
Is Qwen3.8 2.4T A95B free to use?›
There is no free API tier, but the weights are open: you can run it on your own hardware for the cost of compute.
How much does prompt caching save?›
Cache reads cost $0.25 per 1M, 87% below the input price. Writing to the cache costs $2.50 per 1M.
What is a cheaper alternative to Qwen3.8 2.4T A95B?›
Claude Opus 5.5 at high effort costs -15% per task with a score +34% on Humanity's Last Exam.