Qwen 3.8 30B API Pricing — $0.60/M input, $3.00/M output
Alibaba's newest multimodal MoE model: improved agentic coding and reasoning over 3.6, served on Groq LPUs.
Speed
Tokens / sec
690
TTFT
150 ms
Rank
#5 of 31
$ / M ÷ t/s
$0.0017
Measured with 5 runs on a fixed prompt — see the full methodology.
Cost at scale
| Tokens / month | Est. cost (blended 3:1) |
|---|---|
| 100,000 | $0.12 |
| 1,000,000 | $1.20 |
| 10,000,000 | $12.00 |
| 100,000,000 | $120.00 |
How Qwen 3.8 30B compares
Llama 3.1 8B — $0.06/MGPT-OSS 20B — $0.13/MGPT-OSS 120B — $0.26/MLlama 4 Maverick — $0.30/MLlama 3.3 70B — $0.64/MQwen 3.6 27B — $1.20/MQwen 3.7 Plus — $1.10/MGLM-5.1 — $1.00/M
FAQ
Is Qwen 3.8 30B cheaper than Qwen 3.6 27B?
Qwen 3.8 30B costs $1.20/M blended tokens, Qwen 3.6 27B costs $1.20/M — Qwen 3.6 27B is cheaper.
How much does 1 million tokens cost with Qwen 3.8 30B?
At a 3:1 input:output ratio, 1 million blended tokens costs approximately $1.20. Pure input costs $0.60/M; pure output costs $3.00/M.
Who provides Qwen 3.8 30B?
Qwen 3.8 30B is served by Groq. Pricing was last verified on 2026-07-10 against https://groq.com/pricing.