Charm Hyper
Inference hostCharm Hyper runs open and licensed models on its own hardware and sells them by usage, often faster or cheaper than the maker.
Models21
- Cheapest for
- 1 models
- Serves
- 1 model types
Monthly visits—
Models by type
Models on Charm Hyper
| Model | Input | Output | vs official | Context |
|---|---|---|---|---|
| Kimi K3Moonshot AI | $3.27/1M | $16.33/1M from $8.28 at Kenari | +18% | — |
 GLM-5.3Zhipu AI | $1.52/1M | $4.79/1M from $1.10 at NanoGPT | +9% | — |
 Qwen3.8 2.4T A95BAlibaba | $2/1M | $6/1M from $4.11 at APIMart | ±0% | — |
 DeepSeek V4.1 FlashDeepSeek | $0.33/1M | $1.31/1M from $0.03 at UnoRouter | +4267% | — |
| GLM-5.3-FlashZhipu AI | $0.163/1M | $0.544/1M from $0.20 at APIMart | +9% | — |
| Qwen3.8 27BAlibaba | $0.50/1M | $3/1M from $0.50 at EmpirioLabs AI | ±0% | — |
| GLM-5.2Zhipu AI | $1.52/1M | $4.79/1M from $1.80 at Deep Infra | +9% | — |
| DeepSeek V4 Flash Vision ExpDeepSeek | $0.33/1M | $1.31/1M from $0.294 at Deep Infra | +118% | — |
| GPT OSS 120BOpenAI | $0.178/1M | $0.68/1M from $0.14 at Requesty | +386% | — |
 MiniMax-M3MiniMax | $0.327/1M | $1.31/1M from $0.90 at EmpirioLabs AI | +9% | — |
| Qwen3.8 FlashAlibaba | $0.15/1M | $0.47/1M from $0.309 at APIMart | +23% | — |
| Kimi K2.7 Code HighSpeedMoonshot AI | $1.03/1M | $4.36/1M from $2.76 at Kenari | +16% | — |
| DeepSeek V4 Flash 0731DeepSeek | $0.44/1M | $1.32/1M from $0.133 at OpenRouter | +890% | — |
| Qwen3.7 MaxAlibaba | $2.50/1M | $7.50/1M from $3.75 at DevPass (LLM Gateway) | +51% | — |
 Gemma 4 26B A4B ITGoogle | $0.098/1M | $0.334/1M from $0.225 at OpenRouter | -44% | — |
| Qwen3.7 PlusAlibaba | $1.20/1M | $4.80/1M from $0.30 at CrazyRouter | +336% | — |
| DeepSeek V4 ProDeepSeek | $1.44/1M | $4.31/1M from $1.58 at CometAPI | +118% | — |
 Inkling (256K)Thinking Machines | $1.09/1M | $4.41/1M from $1.50 at Venice AI | +194% | — |
| Kimi K2 ThinkingMoonshot AI | $0.60/1M | $2.50/1M from $1.68 at APIMart | +49% | — |
| MiniMax-M2.7MiniMax | $0.484/1M | $1.85/1M from $0.60 at EmpirioLabs AI | +209% | — |
| Qwen3.8 Max PreviewAlibabaCheapest | $2/1M | $6/1M | — | — |
Charm Hyper at a glance
21 models in total: 21 text. The cheapest is Gemma 4 26B A4B IT at $0.334 /1M. Kimi K3 has the largest context window, 1.05M tokens. The most searched is Kimi K3, with about 658K Google searches a month. The newest is DeepSeek V4 Flash Vision Exp, released Oct 17, 2026.
Cheapest
- Gemma 4 26B A4B IT$0.334/1M
- Qwen3.8 Flash$0.47/1M
- GLM-5.3-Flash$0.544/1M
- GPT OSS 120B$0.68/1M
- MiniMax-M3$1.31/1M
Largest context
- Kimi K31.05M
- GLM-5.31.05M
- DeepSeek V4.1 Flash1.05M
- GLM-5.3-Flash1.05M
- DeepSeek V4 Flash Vision Exp1.05M
Most searched
- Kimi K3658K
- GLM-5.3276K
- Qwen3.8 2.4T A95B234K
- DeepSeek V4.1 Flash138K
- GLM-5.3-Flash121K
Newest
- DeepSeek V4 Flash Vision ExpOct 17, 2026
- DeepSeek V4.1 FlashSep 10, 2026
- Qwen3.8 FlashAug 26, 2026
- GLM-5.3Aug 25, 2026
- GLM-5.3-FlashAug 25, 2026