// Alibaba (Qwen)
Qwen3.8-27B
A Alibaba (Qwen) model. 8 prices across 8 channels side by side, each with its source and check date – plus specs, licence, speed and benchmarks.
- Official in/out per 1M
- ¥3 / ¥12
- Lowest (Chutes)
- $0.039 / $1.25
- Context
- 1M
- AA index
- 33.7
Prices checked 2026-10-22Official sources linked on every rowNo estimates
01Specs at a glance
Official values first; gaps filled from OpenRouter’s API where the vendor page is silent.
| Vendor | Alibaba (Qwen) |
|---|---|
| Context window | 1M (1,000,000 tokens) |
| Max output | 236K |
| Input | Text · Image · Video |
| Output | Text |
| Tool calling / JSON | Yes / Yes |
| Weights & licence | Open weights · Apache-2.0 · Hugging Face |
| Parameters | 27.8B (from weight files) |
| Released | Aug 14, 2026 |
| Model IDs | qwen3.8-27b · qwen/qwen3.8-27b · Qwen/Qwen3.8-27B |
02Prices on 8 channels
Per 1M tokens, in the currency each source publishes (1 USD = 6.7 CNY for the blended column). Sorted: official first, then cheapest.
| Channel | Source | Input | Output | Cache hit | Blended $ | Context | Notes |
|---|---|---|---|---|---|---|---|
| Official Vendor’s China site, billed in CNY | Official 10-10 | ¥3 | ¥12 | — | 0.784 | 1M | Batch −50%Beijing region; batch -50% on supported models; explicit cache create 125% / hit 10% of input |
| Chutes · fp8 Via OpenRouter / provider listed there | OpenRouter 10-22 | $0.039 | $1.25 | $0.0039 | 0.342 | 262K | Chutes price as listed on OpenRouter (pass-through; quantization=fp8; discount field=0); not checked against Chutes's own page |
| DeepInfra · bf16 Via OpenRouter / provider listed there | OpenRouter 10-22 | $0.15 | $1.88 | $0.037 | 0.581 | 262K | DeepInfra price as listed on OpenRouter (pass-through; quantization=bf16; discount field=0.25); not checked against DeepInfra's own page |
| Parasail · fp8 Via OpenRouter / provider listed there | OpenRouter 10-22 | $0.24 | $2.2 | $0.05 | 0.73 | 262K | Parasail price as listed on OpenRouter (pass-through; quantization=fp8; discount field=0); not checked against Parasail's own page |
| SiliconFlow Inference platform (own price page) | Platform 10-10 | ¥3 | ¥12 | — | 0.784 | — | SiliconFlow (硅基流动) own pricing page |
| Novita · unknown Via OpenRouter / provider listed there | OpenRouter 10-22 | $0.42 | $3 | $0.085 | 1.06 | 1M | Novita price as listed on OpenRouter (pass-through; quantization=unknown; discount field=0); not checked against Novita's own page |
| Cerebras · fp16 Via OpenRouter / provider listed there | OpenRouter 10-22 | $0.99 | $1.49 | $0.99 | 1.11 | 66K | Cerebras price as listed on OpenRouter (pass-through; quantization=fp16; discount field=0); not checked against Cerebras's own page |
| Groq Inference platform (own price page) | Platform 10-10 | $0.8 | $4 | — | 1.6 | 1M | Groq own pricing page; Groq-reported speed ~450 tok/s |
“Official” = vendor pricing page; “Platform” = the platform’s own pricing page; “OpenRouter” = price listed on OpenRouter, not checked on the platform’s site. Click a source tag to open the original page.
03Where you can use it
Derived only from the price sources above. Country restrictions are set by each vendor – check their terms before you sign up.
- Vendor’s China site, billed in CNY
Official - Via OpenRouter / provider listed there
Chutes · DeepInfra · Parasail · Novita · Cerebras - Inference platform (own price page)
SiliconFlow · Groq - Self-host the open weights
GPU / VRAM sizing and build plans→ - Free option
Qwen3.8-27B→ - Free option
Qwen3.8-27B (Groq)→ - Free option
Qwen 3.8 27B (Cerebras trial)→
04Performance
Speed and AA index from Artificial Analysis public leaderboards (not our own tests); benchmark scores from our benchmark library, best result per benchmark.
- AA indexQwen3.8 27B (Xhigh)
- 33.7
- Output speed (median)
- 40 tok/s
- Time to first tokenQwen3.8 27B (Low): 3.97s
- 3.96s
- AA index per $ (blended)
- 42.98
| Benchmark | Best score | Source | Date |
|---|---|---|---|
| LMArena Text | 1438 | official | 2026-10-08 |
| LMArena WebDev | 1593 | official | 2026-10-08 |
| LMArena Vision | 1243 | official | 2026-10-02 |
05Price changes we recorded
All price changes →From our daily re-checks of the sources above.
- 2026-10-13 · Chutes · cache read0.024 → 0.0039 USD (-83.8%)
- 2026-10-13 · Chutes · output2.2 → 1.25 USD (-43.2%)
- 2026-10-13 · Chutes · input0.24 → 0.039 USD (-83.8%)
06More from Alibaba (Qwen)
FAQ
How much does the Qwen3.8-27B API cost?
On the official pricing page: ¥3 / ¥12 per 1M input / output tokens (checked 2026-10-10). The lowest price we found is $0.039 / $1.25 on Chutes.
What is the context window of Qwen3.8-27B?
1M tokens of context, up to 236K output tokens. Platforms may cap it lower – see the table.
Is Qwen3.8-27B open source?
Yes, the weights are published on Hugging Face (Qwen/Qwen3.8-27B). Licence: Apache-2.0.