AtlasCloud: Models and Prices
Inference platform for text, image and video models, built with the SGLang team.
AtlasCloud served 1.17T tokens of other labs' open-weight models over Oct 3 to Oct 9, 2026 through a large public LLM router, #14 among providers.
Open-weight models
| Model | Family | Share of model | Tokens, 7d | Input $/M | Output $/M | Tokens/s | Uptime 3d | Context |
|---|---|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flash | DeepSeek | 2.7% | 1.02T | $0.30 | $1.20 | 148 | 99.7% | 1.05M |
| DeepSeek V4 Flash 0423 | DeepSeek | 1.3% | 41B | $0.14 | $0.28 | 66 | 98.8% | 1.05M |
| GLM 5.3 Flash | GLM | 0.3% | 30B | $0.15 | $0.50 | 45 | 93.3% | 1.05M |
| DeepSeek V3.2 | DeepSeek | 9.8% | 28B | $0.26 | $0.38 | 28 | 97.6% | 164K |
| DeepSeek V4 Flash 0731 | DeepSeek | 0.3% | 13B | $0.44 | $1.32 | 47 | 99.8% | 1.05M |
| LongCat 2.0 | Other open-weight | 100.0% | 10B | $0.30 | $1.20 | 43 | 99.8% | 1.05M |
| GLM 5.2 | GLM | 0.6% | 9B | $0.938 | $2.95 | 39 | 100.0% | 1.05M |
| Qwen3.6 35B A3B | Qwen | 21.6% | 9B | $0.186 | $1.11 | 29 | 96.8% | 262K |
| MiniMax M3 | MiniMax | 0.3% | 3B | $0.30 | $1.20 | 81 | 98.7% | 524K |
| Kimi K2.5 | Kimi | 6.2% | 950M | $0.49 | $2.50 | 53 | 98.3% | 262K |
| Kimi K2.6 | Kimi | 1.0% | 912M | $0.95 | $4.00 | 56 | 99.0% | 262K |
| GLM 5.3 | GLM | - | - | $1.40 | $4.40 | 45 | 98.8% | 1.05M |
| Hy3 | Hunyuan | - | - | $0.20 | $0.80 | 95 | 100.0% | 262K |
| DeepSeek V4 Pro 0813 | DeepSeek | - | - | $1.32 | $3.96 | 60 | 100.0% | 1.05M |
| DeepSeek V4 Pro 0423 | DeepSeek | - | - | $1.68 | $3.38 | 38 | 99.6% | 1.05M |
| MiMo-V2.5-Pro | MiMo | - | - | $0.435 | $0.87 | - | 5.6% | 1.02M |
| MiniMax M2.7 | MiniMax | - | - | $0.30 | $1.20 | 38.5 | 98.5% | 197K |
| Qwen3.5 397B A17B | Qwen | - | - | $0.55 | $3.50 | 56.5 | 84.0% | 262K |
| GLM 5.1 | GLM | - | - | $1.26 | $3.96 | 32 | 97.0% | 203K |
| MiniMax M2.5 | MiniMax | - | - | $0.295 | $1.20 | 88 | 100.0% | 197K |
| Qwen3.5-35B-A3B | Qwen | - | - | $0.225 | $1.80 | 70 | 100.0% | 262K |
| Qwen3.5-27B | Qwen | - | - | $0.27 | $2.16 | 38 | 99.6% | 262K |
| Qwen3.5-122B-A10B | Qwen | - | - | $0.30 | $2.40 | 85 | 99.7% | 262K |
List prices per million tokens as of October 9, 2026. Share: of all tokens served on that model; a dash where the volume is not reported.
Company
- Website
- atlascloud.ai
- Headquarters
- New York, NY
Sources: SiliconANGLE.
Finance your GPUs
Serving tokens on your own GPUs? Tell us the cluster, the models you serve and the contracts behind them, and we will share it with partner funders that fit.
Prefer email? hello@amcompute.com