Wafer pricing overview
TokenWatch tracks 15 text, 0 image, and 0 video model records for this provider. TokenWatch does not have a provider-wide zero-retention verdict for this page.
Cheapest input price: DeepSeek V4 Flash 0731 at $0.046 per million input tokens and $0.300 per million output tokens. 14 of 15 text models publish a cache-read price; 10 carry a zero-data-retention tag.
Wafer text-model pricing
Text offerings ranked by the Agentic workload mix (2.5% input, 97% cached input, 0.5% output). Change the calculator mix for a workload-specific estimate.
| Org | Provider | Model | Input $/M | Output $/M | Cache $/M | Blended $/M |
|---|---|---|---|---|---|---|
| deepseek | wafer | DeepSeek-V4.1-Flash-1017 | $0.130 | $0.500 | $0 | $0.0057 |
| deepseek | wafer | DeepSeek-V4.1-Flash | $0.190 | $0.650 | $0.010 | $0.018 |
| z-ai | wafer | GLM-5.3-Flash | $0.050 | $0.200 | $0.030 | $0.031 |
| wafer | wafer | DeepSeek-V4-Flash-0731-Fast | $0.130 | $0.230 | $0.030 | $0.033 |
| wafer | wafer | DeepSeek-V4-Flash-0423 | $0.140 | $0.280 | $0.030 | $0.034 |
| deepseek | wafer | DeepSeek: DeepSeek V4 Flash 0731 | $0.046 | $0.300 | $0.045 | $0.046 |
| qwen | wafer | Qwen3.8-27B | $0.060 | $2.30 | $0.050 | $0.062 |
| deepseek | wafer | DeepSeek-V4-Pro | $0.280 | $3.96 | $0.040 | $0.066 |
| deepseek | wafer | DeepSeek: DeepSeek V4 Flash 0731 (fast) | $0.085 | $0.230 | $0.080 | $0.081 |
| deepseek | wafer | DeepSeek: DeepSeek V4 Pro 0813 | $0.115 | $4.30 | $0.110 | $0.131 |
| z-ai | wafer | GLM-5.3 | $0.300 | $3.00 | $0.180 | $0.197 |
| z-ai | wafer | GLM-5.2 | $0.190 | $4.40 | $0.180 | $0.201 |
| deepseek | wafer | DeepSeek: DeepSeek V4 Flash 0423 | $0.600 | $0.170 | $0.200 | $0.210 |
| moonshot | wafer | MoonshotAI: Kimi K3 (fast) | $0.171 | $22.00 | $0.170 | $0.279 |
| moonshot | wafer | Kimi-K3 | $2.00 | $15.00 | $0.250 | $0.367 |
Catalog snapshot 2026-10-29 from direct-provider APIs, OpenRouter endpoint data, and maintained fallbacks. See the sources and calculation rules; verify rates on the provider's official pricing page before committing spend.