All models
Alibaba (Qwen)
Qwen3.5-122B-A10B balancedruns locally
alibaba/qwen3-5-122b-a10b
The Qwen3.5 122B-A10B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. In terms of...
Input / 1M
$0.26
USD per million tokens
Output / 1M
$2.08
USD per million tokens
Est. cost / token
$0.000000715
$0.72 per 1M, 3:1 mix
Context window
262K
max output 66K
Major features
VisionOpen weightsReasoningJSON schemaTools
Task suitability (heuristic 1–5, not a benchmark)
- coding4
- reasoning4
- math4
- summarization4
- extraction4
- classification4
- chat4
- creative4
- translation4
- vision4
- long context3
- agentic4
Run it locally
Will it run on my machine? →| Quantization | Quality | Memory needed (8K ctx) |
|---|---|---|
| Q2_K | poor | 59 GB |
| Q3_K_M | fair | 68.2 GB |
| Q4_K_Mrecommended | good | 84.3 GB |
| Q5_K_M | very good | 97.4 GB |
| Q6_K | excellent | 112 GB |
| Q8_0 | near-lossless | 144 GB |
| F16 | lossless | 267 GB |
Estimates. Mixture-of-experts: memory covers all parameters; only the active ones are used per token, so it runs faster than its size suggests.
Price history
| Recorded | Input | Output | Cached | Source |
|---|---|---|---|---|
| 2026-11-05 03:20 | $0.26 | $2.08 | — | openrouter |
A row is added whenever a sync detects a price change. Vendor pricing source
Pricing (USD per 1M tokens)
- Input
- $0.26
- Output
- $2.08
- Blended (3:1)
- $0.72
Details
- Company
- Alibaba (Qwen)
- Version
- 3.5
- Family
- Qwen
- Released
- 2026-02-25
- Parameters
- 125.1B (10B active)
- Max output
- 66K
- Open weights
- Yes
- Inputs
- image, text, video
- Last synced
- 21 min ago