Qwen3.8 27B
AlibabaOpen weightsApache 2.0Released Aug 14, 2026
Qwen3.8 27B ranks #73 on LMArena (1441.1); official output price $3.00 per 1M tokens.
Key metrics
Rank history
LMArena rank, last 17 days#68 on Sep 23, #73 now; best #68 in this period.
Reasoning effort comparison
Artificial Analysis measurements per reasoning effort of the same model| Effort | Intelligence | Coding | tok/s | TTFT (s) | First answer (s) |
|---|---|---|---|---|---|
| xhighUsed in table | 33.7 | 68.1 | 47 | 1.1 | 44 |
| medium | 27.6 | 56.1 | 51 | 1.2 | 40 |
| low | 26.2 | 58.2 | 55 | 1.2 | 38 |
| non-reasoning | 20.2 | 44.6 | 53 | 1.2 | 1.2 |
Models ranked nearby
Sorted by Arena score| # | Model | Arena | Intel. | Output $ | tok/s |
|---|---|---|---|---|---|
| 48 |  GPT-5.1 | 1442.7 | 24.7 | $10.00 | – |
| 49 |  Claude Sonnet 5 | 1442.1 | 38.2 | $10.00 | – |
| 50 | SStep 5 Preview | 1441.7 | 43.7 | $2.70 | 90 |
| 51 | Qwen3.8 27BThis page | 1441.1 | 33.7 | $3.00 | 47 |
| 52 | GPT-6 Astra | 1440.5 | 52.7 | $50.00 | 48 |
| 53 |  GLM-4.6 | 1439.8 | 14.9 | $2.20 | – |
| 54 |  Hy3 | 1438.8 | 25.3 | $0.56 | 90 |
More from Alibaba
| # | Model | Arena | Intel. | Output $ | tok/s |
|---|---|---|---|---|---|
| 12 | Qwen3.8 Max | 1483.3 | 45.4 | $6.00 | 36 |
| 19 | Qwen3.7 Max | 1475.7 | 29.5 | $7.50 | – |
| 38 | Qwen3.7 Plus | 1454.8 | 25.2 | $1.60 | 55 |
| 43 | Qwen3.6 Max | 1446.9 | 28.4 | $7.80 | – |
| 65 | Qwen3 Max | 1412.8 | 15.6 | $6.00 | – |
| – | Qwen3.8-Flash-Next | – | 39.8 | $0.47 | 53 |
Pricing
- Official API list price: $0.50 input and $3.00 output per 1M tokens (USD).
- At a typical 3:1 input-to-output ratio, the blended cost is about $1.13 per 1M tokens.
- Among 78 models with an output price, it is the #24 cheapest.
- The cheapest OpenRouter route charges $0.43 input / $2.55 output.
Sources & methodology
Qwen3.8 27B (Alibaba): Arena 1441.1 (#73), Intelligence Index 33.7, $0.50 in / $3.00 out per 1M tokens, 1M context. Data updated Oct 9, 2026.
- LMArena — Arena score, 95% confidence interval, votes and rank.
- Artificial Analysis — Intelligence / Coding Index, output speed and latency (medians).
- OpenRouter — Route prices, context length and listing date.
Arena scores use the Bradley-Terry model (not Elo); ± is the half-width of the 95% confidence interval. Latency figures are Artificial Analysis medians; for reasoning models, time to first token and to first answer include the thinking phase. Data syncs daily at 14:20 (UTC+8).
FAQ
What rank is Qwen3.8 27B right now?
Qwen3.8 27B is #51 on our leaderboard of 68 scored models (Arena score 1441.1); its raw LMArena rank is #73.
What does Qwen3.8 27B cost via API?
Input $0.50 / output $3.00 per 1M tokens (official API list price; the OpenRouter quote when no official price is available). At a typical 3:1 input-to-output ratio, the blended cost is about $1.13 per 1M tokens.
How fast is Qwen3.8 27B?
Artificial Analysis measures about 47 output tokens per second, 1.1 s to the first token and 44 s to the first answer token; it is a reasoning model, so latency includes thinking time.
How close is Qwen3.8 27B to the models ranked next to it?
It is 0.6 points behind Step 5 Preview (1441.7) and 0.6 points ahead of GPT-6 Astra (1440.5); gaps smaller than the confidence interval (±5.3) are not significant.
Is Qwen3.8 27B open source? Can it be self-hosted?
Qwen3.8 27B is an open-weights model (license: Apache 2.0); you can download the weights and self-host.
What is the context length of Qwen3.8 27B?
1M (1,000,000 tokens, as listed on OpenRouter).