Deep Infra API Pricing
🌍 GlobalAPI price snapshot
Current Deep Infra models and token prices
ComputeUnion currently tracks 17 Deep Infra models, with input pricing from $0.06 per 1M tokens. Each model detail page keeps this platform's quote separate from other available official or channel quotes so the official API versus relay cost gap can be checked directly.
Payment Methods
Deep Infra is an AI API relay service. ComputeUnion tracks 22 price records for this platform (22 auto-scraped). Browse the categories below to compare Deep Infra pricing across providers.
Model Pricing on Deep Infra
Prices per 1M tokens
LLM17 models
| Model | Input | Output | Context | Updated |
|---|---|---|---|---|
| DeepSeek V3.2 deepseek-v3-2 | $0.2600 | $0.3800 | 160K | Updated 1h ago |
| DeepSeek V4 deepseek-v4 | $0.0600 | $0.1800 | 1M | Updated 1h ago |
| DeepSeek V4 Flash deepseek-v4-flash | $0.0900 | $0.1800 | 1M | Updated 1h ago |
| DeepSeek V4 Pro deepseek-v4-pro | $1.3000 | $2.6000 | 1M | Updated 1h ago |
| Gemma 4 27B gemma-4-27b | $0.0700 | $0.3400 | 256K | Updated 1h ago |
| Gemma 4 31B gemma-4-31b | $0.1300 | $0.3800 | 256K | Updated 1h ago |
| GLM-5 glm-5 | $0.1500 | $0.5000 | 1M | Updated 1h ago |
| GLM-5.1 glm-5-1 | $1.0500 | $3.5000 | 198K | Updated 1h ago |
| GLM-5.2 glm-5-2 | $0.7500 | $2.4000 | 1M | Updated 1h ago |
| Kimi kimi | $0.6800 | $3.4000 | 256K | Updated 1h ago |
| Kimi K2.6 kimi-k2-6 | $0.7500 | $3.5000 | 256K | Updated 1h ago |
| MiMo V2.5 Pro mimo-v2-5-pro | $1.0000 | $3.0000 | 1M | Updated 1h ago |
| NVIDIA Nemotron nvidia-nemotron | $0.0850 | $0.4000 | 256K | Updated 1h ago |
| Qwen qwen | $1.2000 | $6.0000 | 250K | Updated 1h ago |
| Qwen3.5 397B A17B qwen3-5-397b-a17b | $0.4500 | $3.0000 | 256K | Updated 1h ago |
| Qwen3.6 35B A3B qwen3-6-35b-a3b | $0.1000 | $0.9500 | 256K | Updated 1h ago |
| Qwen3 Max qwen3-max | $1.2000 | $6.0000 | 250K | Updated 1h ago |
Sources and price verification method
This page lists Deep Infra price records by model and distinguishes live-API, independently checked public-page, and maintained records. Platform identity, model availability, and pricing are separate evidence layers: official sources confirm vendor or model information, while ComputeUnion price records show currently comparable quotes. Relay model IDs, terms, refund policies, and regional availability should still be checked with the platform before purchase.
No vendor source is currently registered for automatic official-status decisions; this page only presents reviewable platform and pricing records already in the database.
FAQ
What is Deep Infra?
Deep Infra is an AI API relay service aggregating multi-provider access through a unified endpoint. ComputeUnion currently tracks 22 price records for this platform (22 auto-scraped).
How does Deep Infra pricing compare to official providers?
Deep Infra quotes may differ from the official API, as may model IDs, regions, and service terms. Use the model links below to compare currently tracked quotes and verify current terms before purchase.
Is Deep Infra accessible from China?
Deep Infra is an international service — direct access from China mainland may be restricted.
Related pages