DeepInfra API pricing
47 models · USD per 1M tokens · verified 2026-10-23
DeepInfra serves 47 of the models we track, 43 of them open-weights. The cheapest is Mistral Small 3 at $0.05 input and $0.08 output per million tokens. For 40 of these models, DeepInfra is the cheapest tracked host.
// models on DeepInfra
| model | lab | in $/1M | out $/1M | ctx | vs best price |
|---|---|---|---|---|---|
| Mistral Small 3 BEST | Mistral | $0.05 | $0.08 | 32.8K | best |
| GPT OSS 20B BEST | OpenAI | $0.03 | $0.14 | 131.1K | best |
| GPT OSS 120B BEST | OpenAI | $0.037 | $0.17 | 131.1K | best |
| Gemma 3 12B IT | $0.05 | $0.15 | 131.1K | +20% | |
| DeepSeek V4 Flash 0731 BEST | DeepSeek | $0.06 | $0.18 | 1.05M | best |
| Gemma 3 27B IT BEST | $0.08 | $0.16 | 131.1K | best | |
| Qwen3.5 9B BEST | Alibaba (Qwen) | $0.10 | $0.15 | 262.1K | best |
| DeepSeek V4 Flash BEST | DeepSeek | $0.09 | $0.18 | 1.05M | best |
| Gemma 4 26B A4B IT BEST | $0.07 | $0.34 | 262.1K | best | |
| Llama 4 Scout 17B BEST | Meta | $0.10 | $0.30 | 327.7K | best |
| Llama 3.3 70B Turbo BEST | Meta | $0.10 | $0.32 | 131.1K | best |
| Llama Guard 4 12B BEST | Meta | $0.18 | $0.18 | 163.8K | best |
| Qwen3.8 Flash BEST | Alibaba (Qwen) | $0.113 | $0.382 | 1M | −21.6% |
| Qwen3 235B-A22B Instruct 2507 BEST | Alibaba (Qwen) | $0.09 | $0.55 | 262.1K | best |
| Qwen3 30B A3b fp8 | Alibaba (Qwen) | $0.12 | $0.50 | 32.8K | +76.4% |
| GLM-5.3-FlashX BEST | Z.AI (Zhipu) | $0.15 | $0.50 | 1M | −59.7% |
| Gemma 4 31B IT | $0.20 | $0.40 | 262.1K | +22% | |
| Qwen3 VL 30B A3B Instruct BEST | Alibaba (Qwen) | $0.15 | $0.60 | 262.1K | best |
| DeepSeek-V3.2 BEST | DeepSeek | $0.26 | $0.38 | 163.8K | best |
| DeepSeek V4.1 Flash BEST | DeepSeek | $0.20 | $0.60 | 1.05M | best |
| Qwen3.6 35B-A3B BEST | Alibaba (Qwen) | $0.10 | $0.95 | 262.1K | −63% |
| Qwen3.5 35B-A3B BEST | Alibaba (Qwen) | $0.14 | $1.00 | 262.1K | −48.4% |
| Qwen3 VL 235B A22B Instruct BEST | Alibaba (Qwen) | $0.20 | $0.88 | 262.1K | best |
| DeepSeek-V3.1 BEST | DeepSeek | $0.25 | $0.95 | 163.8K | best |
| DeepSeek-V3 BEST | DeepSeek | $0.32 | $0.89 | 163.8K | best |
| MiniMax-M3 BEST | MiniMax | $0.28 | $1.10 | 1M | −7.6% |
| DeepSeek V4 Flash Vision Exp BEST | DeepSeek | $0.44 | $1.32 | 1.05M | best |
| GLM-4.7 BEST | Z.AI (Zhipu) | $0.40 | $1.75 | 204.8K | −26.2% |
| Qwen3.8 27B BEST | Alibaba (Qwen) | $0.20 | $2.50 | 262.1K | best |
| Qwen3.5 122B-A10B BEST | Alibaba (Qwen) | $0.29 | $2.40 | 262.1K | −25.7% |
| Qwen3.5 27B | Alibaba (Qwen) | $0.26 | $2.60 | 262.1K | +2.4% |
| GLM-4.6 BEST | Z.AI (Zhipu) | $0.50 | $2.00 | 204.8K | −12.5% |
| DeepSeek-R1-0528 BEST | DeepSeek | $0.50 | $2.15 | 163.8K | best |
| GLM-5 BEST | Z.AI (Zhipu) | $0.60 | $2.08 | 204.8K | −37.4% |
| Qwen3.6 27B BEST | Alibaba (Qwen) | $0.32 | $3.20 | 262.1K | −23% |
| Qwen3.5 397B-A17B BEST | Alibaba (Qwen) | $0.45 | $3.00 | 262.1K | −19.4% |
| GLM-5.2 BEST | Z.AI (Zhipu) | $0.75 | $2.40 | 1M | −45.9% |
| Kimi K2.7 Code HighSpeed BEST | Moonshot AI | $0.68 | $3.40 | 262.1K | −60.3% |
| Kimi K2.6 BEST | Moonshot AI | $0.75 | $3.50 | 262.1K | −16.1% |
| DeepSeek V4 Pro BEST | DeepSeek | $1.30 | $2.60 | 1.05M | best |
| DeepSeek V4 Pro (0813) BEST | DeepSeek | $1.30 | $2.60 | 1.05M | +64.1% |
| GLM-5.1 BEST | Z.AI (Zhipu) | $1.05 | $3.50 | 200K | −22.7% |
| GLM-5.3 | Z.AI (Zhipu) | $0.90 | $4.00 | 1M | −22.1% |
| Qwen3.8 Max BEST | Alibaba (Qwen) | $1.65 | $4.951 | 1M | −17.5% |
| Qwen3.7 Max | Alibaba (Qwen) | $2.50 | $7.50 | 1M | best |
| Kimi K3 | Moonshot AI | $2.85 | $14.25 | 1.05M | −5% |
| Claude Fable 5 BEST | Anthropic | $10.00 | $50.00 | 1M | best |
// how to read this page
Prices are DeepInfra's base-tier list prices per million tokens. "vs best price" compares DeepInfra's blended price (3 × input + output) / 4 with the lowest price we track for the same model, whether it comes from the lab's own API or another host. Open each model for its full price history and every host that serves it.