Qwen3.8 27B cut across 2 hosts, by up to 2% at Reka (input)
Published 26 September 2026
Price move, per 1M tokens
- Reka · input↓ 2%$0.092 → $0.090
- Reka · outputunchanged$4.40 → $4.40
- Reka · cache readunchanged$0.085 → $0.085
- Wafer · input↓ 2%$0.093 → $0.091
- Wafer · outputunchanged$4.40 → $4.40
- Wafer · cache readunchanged$0.085 → $0.085
| host | rate | was | now | change |
|---|---|---|---|---|
| Reka | input | $0.092 | $0.090 | ↓ 2% |
| Reka | output | $4.40 | $4.40 | — |
| Reka | cache read | $0.085 | $0.085 | — |
| Wafer | input | $0.093 | $0.091 | ↓ 2% |
| Wafer | output | $4.40 | $4.40 | — |
| Wafer | cache read | $0.085 | $0.085 | — |
Reka input: ↓ 2%Wafer input: ↓ 2%
If you buy
a slightly lower input rate at Wafer makes this a good moment to re-check your usage mix and confirm the new rate applies to your next top-up.
The model
Qwen3.8 27Bopen the model →
Open weights27.8B parameters262K context20 hosts
intelligence63rd of 168 writing93rd of 168 coding45th of 168 agents33rd of 55
Qwen3.8 27B is a downloadable model you can run yourself, and its measured strength is maths and reasoning. Agentic recovery and tool use sit near the bottom of the field, so it suits a solver rather than an agent that has to get itself back on track.
Source
Read on openrouter.ai · we published this on 26 September 2026.
read the original at openrouter.ai ↗