Inference APIs is a third-party gateway that routes to many upstream models through one API. Gateway quotes are often among the lowest on the site, but availability, rate limits and routing can vary between models, so a low headline price is worth verifying against your own workload.
It lists 11 models on this page. Coverage spans text and chat (11). The lowest-priced model hosted here is GPT OSS 20B.
| Cheapest here? | |||||
|---|---|---|---|---|---|
| GPT OSS 20Bopenai/gpt-oss-20b4.3× pricier | $0.04 | $0.19 | — | 131,072 | 4.3× pricier |
| Qwen3.8 FlashQwen/Qwen3.8-Flash1.4× pricier | $0.15 | $0.50 | $0.02 | 1,000,000 | 1.4× pricier |
| DeepSeek V4 Flashdeepseek-ai/DeepSeek-V4-Flash3.0× pricier | $0.19 | $0.38 | $0.04 | 1,048,576 | 3.0× pricier |
| Llama-3.3-70B-Instructmeta-llama/Llama-3.3-70B-Instruct-Turbo2.9× pricier | $0.20 | $0.50 | — | 131,072 | 2.9× pricier |
| GLM-5.3-Flashzai-org/GLM-5.3-Flash6.4× pricier | $0.20 | $0.66 | $0.04 | 1,000,000 | 6.4× pricier |
| GPT OSS 120Bopenai/gpt-oss-120b6.0× pricier | $0.20 | $0.80 | — | 131,072 | 6.0× pricier |
| MiniMax-M3MiniMaxAI/MiniMax-M31.8× pricier | $0.39 | $1.56 | $0.08 | 524,288 | 1.8× pricier |
| DeepSeek V4.1 Flashdeepseek-ai/DeepSeek-V4.1-Flash13.3× pricier | $0.40 | $1.60 | $0.01 | 1,048,576 | 13.3× pricier |
| Kimi K2.7 Codemoonshotai/Kimi-K2.7-Code3.7× pricier | $0.89 | $4.42 | $0.18 | 262,144 | 3.7× pricier |
| DeepSeek V4 Prodeepseek-ai/DeepSeek-V4-Pro5.9× pricier | $1.72 | $5.15 | $0.17 | 1,048,576 | 5.9× pricier |
| GLM-5.3zai-org/GLM-5.38.1× pricier | $1.82 | $5.72 | $0.34 | 1,000,000 | 8.1× pricier |