Qwen3 235B A22B
MoE workstation235B total, 22B active per forward pass. Q4_K_M fits in ~128GB (DGX Spark territory).
AI-generated content marks
This model embeds text watermarks in generated text and adds C2PA provenance metadata to supported files such as .png, .jpg, and .svg. Marks can be lost through editing, screenshots, or format conversion, so their absence does not prove a file is human-made.
- 235.0B
- 128k
- apache 2.0
- 🇨🇳 China
- Apr 2025
Scores
Score per dollar
187 pts per $/M input
general_score (85) divided by cheapest input price ($0.46/M). Higher is better value. See live pricing.
Related models
Guides covering Qwen3 235B A22B
Save your hardware and every model page answers the real question: will it run on your machine, and how fast?
Join free - save your rig →Run it locally
Per-quant memory needs and a static "can you run it?" reference - no rig entry required
The reference hardware
22 reference configs, drawn in-house. Scroll for more.
Can you run it? - reference rigs
| Rig | Q2_K | Q4_K_M |
|---|---|---|
| NVIDIA Jetson Orin NX 16GB | no -> cloud | no -> cloud |
| Jetson AGX Orin 64GB | no -> cloud | no -> cloud |
| Single GTX 1080 Ti (11GB) | no -> cloud | no -> cloud |
| Single RTX 4090 (24GB) | no -> cloud | no -> cloud |
| Single RTX 5090 (32GB) | no -> cloud | no -> cloud |
| 2x RTX 3090 (48GB) | offload | no -> cloud |
| 2x RTX 5090 (64GB) | offload | no -> cloud |
| 4x H100 80GB (320GB) | fast 1156.3t/s | fast 629.4t/s |
| NVIDIA DGX Station 748GB | fast 690.3t/s | fast 375.7t/s |
| 8x RTX 3090 rack (192GB) | fast 646.3t/s | fast 351.8t/s |
| 4x RTX 5090 (128GB) | fast 618.5t/s | tight |
| AMD Instinct MI300X (192GB) | fast 459.5t/s | fast 250.1t/s |
| 4x RTX 4090 (96GB) | fast 347.9t/s | offload |
| RTX PRO 6000 Blackwell (96GB) | fast 154.6t/s | offload |
| Mac Studio M4 Ultra 192GB | fast 102.8t/s | fast 56.0t/s |
| Mac Studio M4 Ultra 512GB | fast 102.8t/s | fast 56.0t/s |
| MacBook Pro M5 Max 128GB | fast 57.8t/s | tight |
| Dual EPYC 9004 + 768GB DDR5-4800 | fast 39.8t/s | fast 21.6t/s |
| DGX Spark 128GB unified | fast 23.6t/s | tight |
| Ryzen AI Max+ 395 128GB | fast 22.1t/s | tight |
| Epyc + 512GB DDR4-3200 + 2x RTX 3090 | ok 17.7t/s | ok 9.6t/s |
| Epyc + 512GB DDR4-2400 + 2x RTX 3090 | ok 13.3t/s | slow 7.2t/s |
Fit tiers use the same will-it-run logic as the rig finder. For comfortable fits, the badge reflects decode speed: fast >=20 t/s, ok 8-20 t/s, slow <8 t/s. t/s is a bandwidth estimate, not a measured benchmark.
Download options
Or run it in the cloud
Live per-provider pricing, throughput and uptime - refreshed about 9 hours ago via OpenRouter. Click a column to sort.
| Provider | Type | Input $/M | Output $/M | Cache $/M | Tok/s | Latency | Uptime | Value |
|---|---|---|---|---|---|---|---|---|
|
Alibaba
|
API | 0.46 | 1.82 | - | - | - | 99.30% | best uptime |
Default order: throughput among 95%+ uptime providers, then latency; subscriptions last. Sort by any column. Subscription rows show $/mo in the Value column - per-token columns are "-". Affiliate links are marked sponsored / nofollow. Confirm current pricing on the provider's site before committing.
Detailed API pricing page + JSON endpoint →
See who runs Alibaba in production →
Inference cost over time
Data accumulates from the first daily sync - longer ranges populate over time. Prices come from OpenRouter snapshots, not a historical API.