Cheapest LLM APIs
The lowest-cost large language model APIs we track, ranked by combined input and output price per million tokens. Most of the cheapest options are open-weight models (DeepSeek, Qwen, GLM, MiniMax) served by competing providers — which is exactly why prices keep falling.
Ranking 50 options · updated June 30, 2026
| # | Model | Provider | Input /1M | Output /1M |
|---|---|---|---|---|
| 1 | Llama 3.1 8B Instruct open | Novita AI | $0.020 | $0.050 |
| 2 | gpt-oss-20b open | Hyperbolic | $0.040 | $0.040 |
| 3 | Gemma 3 4B open | OpenRouter | $0.050 | $0.100 |
| 4 | Gemma 3 4B open | Together AI | $0.060 | $0.120 |
| 5 | Gemma 3 12B open | OpenRouter | $0.050 | $0.150 |
| 6 | Gemma 3 4B open | Fireworks AI | $0.100 | $0.100 |
| 7 | Gemma 3 27B open | OpenRouter | $0.080 | $0.160 |
| 8 | gpt-oss-20b open | Parasail | $0.040 | $0.200 |
| 9 | Qwen 3.5 9B open | DeepInfra | $0.100 | $0.150 |
| 10 | Qwen 3.5 9B open | OpenRouter | $0.100 | $0.150 |
| 11 | DeepSeek V4 Flash open | OpenRouter | $0.090 | $0.180 |
| 12 | Mistral Small 3.2 24B open | OpenRouter | $0.075 | $0.200 |
| 13 | Ministral 3 8B open | OpenRouter | $0.150 | $0.150 |
| 14 | Nova Lite | OpenRouter | $0.060 | $0.240 |
| 15 | Ministral 3 8B open | Mistral AI | $0.150 | $0.150 |
| 16 | DeepSeek V4 Flash open | DeepInfra | $0.100 | $0.200 |
| 17 | Gemma 3 27B open | Novita AI | $0.119 | $0.200 |
| 18 | Gemma 3 27B open | Venice AI | $0.120 | $0.200 |
| 19 | Mistral Small 4 open | Venice AI | $0.090 | $0.250 |
| 20 | Qwen3 14B open | OpenRouter | $0.100 | $0.240 |
| 21 | Qwen3 Coder 30B A3B Instruct open | OpenRouter | $0.070 | $0.270 |
| 22 | Qwen 3 32B open | OpenRouter | $0.080 | $0.280 |
| 23 | GPT OSS 120B open | Venice AI | $0.070 | $0.300 |
| 24 | gpt-oss-20b open | Fireworks AI | $0.070 | $0.300 |
| 25 | Qwen 3 8B open | Cloudflare Workers AI | $0.051 | $0.335 |
| 26 | Mistral Small 4 open | Parasail | $0.090 | $0.300 |
| 27 | Gemma 4 26B A4B open | OpenRouter | $0.060 | $0.330 |
| 28 | Llama 4 Scout open | OpenRouter | $0.100 | $0.300 |
| 29 | Gemma 3 12B open | Fireworks AI | $0.200 | $0.200 |
| 30 | Qwen 3 8B open | Fireworks AI | $0.200 | $0.200 |
| 31 | Ministral 3 8B open | Fireworks AI | $0.200 | $0.200 |
| 32 | Qwen 3 32B open | Nebius | $0.100 | $0.300 |
| 33 | DeepSeek V4 Flash open | SiliconFlow | $0.130 | $0.280 |
| 34 | Qwen 3 8B open | Together AI | $0.170 | $0.250 |
| 35 | Llama 3.3 70B open | OpenRouter | $0.100 | $0.320 |
| 36 | DeepSeek V3.2 open | DeepSeek | $0.140 | $0.280 |
| 37 | DeepSeek R1 open | DeepSeek | $0.140 | $0.280 |
| 38 | DeepSeek V4 Flash open | Novita AI | $0.140 | $0.280 |
| 39 | DeepSeek V4 Flash open | Fireworks AI | $0.140 | $0.280 |
| 40 | DeepSeek V4 Flash open | Venice AI | $0.140 | $0.280 |
| 41 | DeepSeek V4 Flash open | Parasail | $0.140 | $0.280 |
| 42 | DeepSeek V4 Flash open | Atlas Cloud | $0.140 | $0.280 |
| 43 | DeepSeek V4 Flash open | DeepSeek | $0.140 | $0.280 |
| 44 | Llama 3.1 8B Instruct open | Cloudflare Workers AI | $0.045 | $0.384 |
| 45 | GPT-5 Nano | Replicate | $0.050 | $0.400 |
| 46 | GPT-5 Nano | OpenRouter | $0.050 | $0.400 |
| 47 | GPT-5 Nano | OpenAI | $0.050 | $0.400 |
| 48 | Llama 4 Scout open | Groq | $0.110 | $0.340 |
| 49 | Qwen 3.5 9B open | Alibaba Cloud | $0.050 | $0.400 |
| 50 | GLM 4.7 | DeepInfra | $0.060 | $0.400 |