Maximum value -- models that are cheap AND smart. Filtered to exclude low-quality options (intelligence score > 20). Sorted by input price per million tokens.
GPT-5 nano (high) offers the best value at just $0.05/1M tokens while maintaining strong benchmark quality.
Ranked by input cost per million tokens, filtered to models scoring above 20 on intelligence to exclude low-quality options.
| # | Model | Provider | Price/1M | Intelligence | Speed | Context | |
|---|---|---|---|---|---|---|---|
| 1 | GPT-5 nano (high) | OpenAI | $0.05 | 26.8 | 132 tok/s | N/A | |
| 2 | gpt-oss-20B (high) | OpenAI | $0.06 | 24.5 | 281 tok/s | N/A | |
| 3 |
Based on our benchmark rankings, GPT-5 nano (high) is currently the top-ranked model for budget-conscious projects. See the full rankings above for alternatives.
GPT-5 nano (high) costs $0.05/1M input tokens. For 100 requests per day at 2,000 tokens each, that's approximately $0.30/month.
Currently, there are no free models that rank highly for budget-conscious projects. Check our free models page for the best zero-cost options.
Take our 30-second quiz and get a personalized recommendation.
Take the Quiz| NVIDIA |
| $0.06 |
| 24.3 |
| 176 tok/s |
| N/A |
| 4 | Z | GLM-4.7-Flash (Reasoning) | Z AI | $0.07 | 30.1 | 86 tok/s | N/A |
| 5 | X | MiMo-V2-Flash (Feb 2026) | Xiaomi | $0.10 | 41.5 | 128 tok/s | N/A |
| 6 | A | Qwen3.5 9B (Reasoning) | Alibaba | $0.10 | 32.4 | 49 tok/s | N/A |
| 7 | S | Step 3.5 Flash | StepFun | $0.10 | 37.8 | 85 tok/s | N/A |
| 8 | Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning) | $0.10 | 21.6 | 389 tok/s | N/A |
| 9 | gpt-oss-120B (high) | OpenAI | $0.15 | 33.3 | 254 tok/s | N/A |
| 10 | M | Mistral Small 4 (Reasoning) | Mistral | $0.15 | 26.9 | 0 tok/s | N/A |
| 11 | Z | GLM-4.5-Air | Z AI | $0.20 | 23.2 | 104 tok/s | N/A |
| 12 | X | Grok 4 Fast (Reasoning) | xAI | $0.20 | 35.1 | 128 tok/s | N/A |
| 13 | X | Grok Code Fast 1 | xAI | $0.20 | 28.7 | 195 tok/s | 256K |
| 14 | A | Qwen3 30B A3B 2507 (Reasoning) | Alibaba | $0.20 | 22.4 | 140 tok/s | N/A |
| 15 | GPT-5.4 nano (xhigh) | OpenAI | $0.20 | 44.4 | 216 tok/s | N/A |
| 16 | X | Grok 4.1 Fast (Reasoning) | xAI | $0.20 | 38.6 | 160 tok/s | N/A |
| 17 | B | Seed-OSS-36B-Instruct | ByteDance Seed | $0.21 | 25.2 | 42 tok/s | N/A |
| 18 | GPT-5 mini (high) | OpenAI | $0.25 | 41.2 | 76 tok/s | N/A |
| 19 | GPT-5 mini (minimal) | OpenAI | $0.25 | 20.7 | 76 tok/s | N/A |
| 20 | Gemini 3.1 Flash-Lite Preview | $0.25 | 33.5 | 208 tok/s | N/A |
| 21 | GPT-5.1 Codex mini (high) | OpenAI | $0.25 | 38.6 | 176 tok/s | N/A |
| 22 | A | Qwen3.5 35B A3B (Reasoning) | Alibaba | $0.25 | 37.1 | 130 tok/s | N/A |
| 23 | I | Mercury 2 | Inception | $0.25 | 32.8 | 894 tok/s | 128K |
| 24 | DeepSeek V3.2 Exp (Reasoning) | DeepSeek | $0.28 | 32.9 | 32 tok/s | N/A |
| 25 | DeepSeek V3.2 (Reasoning) | DeepSeek | $0.28 | 41.7 | 32 tok/s | N/A |
| 26 | MiniMax-M2.7 | MiniMax | $0.30 | 49.6 | 45 tok/s | N/A |
| 27 | Gemini 2.5 Flash (Reasoning) | $0.30 | 27.0 | 213 tok/s | N/A |
| 28 | K | KAT-Coder-Pro V1 | KwaiKAT | $0.30 | 36.0 | 40 tok/s | N/A |
| 29 | MiniMax-M2.5 | MiniMax | $0.30 | 41.9 | 52 tok/s | N/A |
| 30 | A | Nova 2.0 Omni (medium) | Amazon | $0.30 | 28.0 | 0 tok/s | N/A |
| 31 | X | Grok 3 mini Reasoning (high) | xAI | $0.30 | 32.1 | 198 tok/s | N/A |
| 32 | NVIDIA Nemotron 3 Super 120B A12B (Reasoning) | NVIDIA | $0.30 | 36.0 | 365 tok/s | N/A |
| 33 | MiniMax-M2.1 | MiniMax | $0.30 | 39.4 | 51 tok/s | N/A |
| 34 | MiniMax-M2 | MiniMax | $0.30 | 36.1 | 52 tok/s | N/A |
| 35 | Z | GLM-4.6V (Reasoning) | Z AI | $0.30 | 23.4 | 28 tok/s | N/A |
| 36 | A | Qwen3.5 27B (Reasoning) | Alibaba | $0.30 | 42.1 | 91 tok/s | N/A |
| 37 | A | Nova 2.0 Lite (medium) | Amazon | $0.30 | 29.7 | 235 tok/s | N/A |
| 38 | A | Qwen3 Coder Next | Alibaba | $0.35 | 28.3 | 161 tok/s | N/A |
| 39 | GPT-4.1 mini | OpenAI | $0.40 | 22.9 | 70 tok/s | 1.0M |
| 40 | DeepSeek V3.1 Terminus (Reasoning) | DeepSeek | $0.40 | 33.9 | 0 tok/s | N/A |
| 41 | M | Mistral Medium 3.1 | Mistral | $0.40 | 21.3 | 78 tok/s | N/A |
| 42 | A | Qwen3.5 122B A10B (Reasoning) | Alibaba | $0.40 | 41.6 | 133 tok/s | N/A |
| 43 | Z | GLM-4.5 (Reasoning) | Z AI | $0.49 | 26.4 | 41 tok/s | N/A |
| 44 | A | Qwen3 Next 80B A3B (Reasoning) | Alibaba | $0.50 | 26.7 | 142 tok/s | N/A |
| 45 | A | Qwen3 Next 80B A3B Instruct | Alibaba | $0.50 | 20.1 | 141 tok/s | N/A |
| 46 | M | Mistral Large 3 | Mistral | $0.50 | 22.8 | 58 tok/s | N/A |
| 47 | Gemini 3 Flash Preview (Reasoning) | $0.50 | 46.4 | 192 tok/s | N/A |
| 48 | MiniMax M1 80k | MiniMax | $0.55 | 24.4 | 0 tok/s | N/A |
| 49 | DeepSeek V3.1 (Non-reasoning) | DeepSeek | $0.56 | 28.1 | 0 tok/s | N/A |
| 50 | K | Kimi K2 | Kimi | $0.57 | 26.3 | 36 tok/s | N/A |