Intelligence Per Dollar
Benchmark points per blended dollar across every credibly ranked model. Scores are min-max normalized across 47 models with 4+ independent benchmarks; cost assumes a typical 3:1 input:output token mix. Leader = 100.
This is not a quality ranking. #1 here means the most benchmark points per dollar, so a mid-scoring budget model will outrank a frontier model that costs 100x more. Check the Score and % of top score columns for raw capability, and use the task rankings when output quality is what compounds in your workflow.
Value leaderboard
| # | Model | Score | % of top score | In / Out per 1M | Blended $/1M | Value |
|---|---|---|---|---|---|---|
| 1 | OpenAI: gpt-oss-120b BEST VALUEopenai/gpt-oss-120b | 51.8 | 52% | $0.04 / $0.18 | $0.07 | 100.0 |
| 2 | OpenAI: gpt-oss-20bopenai/gpt-oss-20b | 37.3 | 37% | $0.03 / $0.14 | $0.06 | 94.2 |
| 3 | DeepSeek: DeepSeek V4 Flashdeepseek/deepseek-v4-flash | 80.4 | 81% | $0.10 / $0.20 | $0.12 | 94.1 |
| 4 | Google: Gemma 4 26B A4B google/gemma-4-26b-a4b-it | 54.6 | 55% | $0.06 / $0.33 | $0.13 | 61.4 |
| 5 | Google: Gemma 4 31Bgoogle/gemma-4-31b-it | 68.5 | 69% | $0.12 / $0.35 | $0.18 | 55.3 |
| 6 | Z.ai: GLM 4.7 Flashz-ai/glm-4.7-flash | 54.1 | 54% | $0.06 / $0.40 | $0.15 | 53.5 |
| 7 | DeepSeek: DeepSeek V3deepseek/deepseek-chat | 84.1 | 84% | $0.20 / $0.80 | $0.35 | 34.4 |
| 8 | Google: Gemma 3 4Bgoogle/gemma-3-4b-it | 12.7 | 13% | $0.05 / $0.10 | $0.06 | 29.1 |
| 9 | Google: Gemma 3 27Bgoogle/gemma-3-27b-it | 17.0 | 17% | $0.08 / $0.16 | $0.10 | 24.4 |
| 10 | DeepSeek: DeepSeek V4 Prodeepseek/deepseek-v4-pro | 88.1 | 88% | $0.43 / $0.87 | $0.54 | 23.2 |
| 11 | Qwen: Qwen3 8Bqwen/qwen3-8b | 18.2 | 18% | $0.05 / $0.40 | $0.14 | 19.0 |
| 12 | MoonshotAI: Kimi K2.5moonshotai/kimi-k2.5 | 80.2 | 80% | $0.35 / $1.89 | $0.73 | 15.6 |
| 13 | Z.ai: GLM 4.7z-ai/glm-4.7 | 73.7 | 74% | $0.40 / $1.75 | $0.74 | 14.3 |
| 14 | OpenAI: GPT-4.1 Nanoopenai/gpt-4.1-nano | 16.7 | 17% | $0.10 / $0.40 | $0.18 | 13.7 |
| 15 | Z.ai: GLM 5z-ai/glm-5 | 85.0 | 85% | $0.60 / $1.92 | $0.93 | 13.1 |
| 16 | Qwen: Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b | 78.3 | 79% | $0.39 / $2.34 | $0.88 | 12.8 |
| 17 | Meta: Llama 4 Maverickmeta-llama/llama-4-maverick | 18.0 | 18% | $0.15 / $0.60 | $0.26 | 9.8 |
| 18 | MoonshotAI: Kimi K2.6moonshotai/kimi-k2.6 | 88.9 | 89% | $0.67 / $3.39 | $1.35 | 9.4 |
| 19 | Z.ai: GLM 5.1z-ai/glm-5.1 | 85.7 | 86% | $0.98 / $3.08 | $1.50 | 8.2 |
| 20 | OpenAI: GPT-4.1 Miniopenai/gpt-4.1-mini | 39.2 | 39% | $0.40 / $1.60 | $0.70 | 8.0 |
| 21 | Google: Gemini 2.5 Flashgoogle/gemini-2.5-flash | 42.2 | 42% | $0.30 / $2.50 | $0.85 | 7.1 |
| 22 | xAI: Grok 4.20x-ai/grok-4.20 | 74.7 | 75% | $1.25 / $2.50 | $1.56 | 6.9 |
| 23 | Z.ai: GLM 4.5z-ai/glm-4.5 | 46.6 | 47% | $0.60 / $2.20 | $1.00 | 6.7 |
| 24 | Anthropic: Claude 3 Haikuanthropic/claude-3-haiku | 16.9 | 17% | $0.25 / $1.25 | $0.50 | 4.8 |
| 25 | OpenAI: o4 Miniopenai/o4-mini | 61.6 | 62% | $1.10 / $4.40 | $1.93 | 4.6 |
| 26 | OpenAI: GPT-5openai/gpt-5 | 82.0 | 82% | $1.25 / $10.00 | $3.44 | 3.4 |
| 27 | OpenAI: o3 Mini Highopenai/o3-mini-high | 44.2 | 44% | $1.10 / $4.40 | $1.93 | 3.3 |
| 28 | OpenAI: o3openai/o3 | 74.2 | 74% | $2.00 / $8.00 | $3.50 | 3.0 |
| 29 | Google: Gemini 3.1 Pro Previewgoogle/gemini-3.1-pro-preview | 92.4 | 93% | $2.00 / $12.00 | $4.50 | 2.9 |
| 30 | OpenAI: GPT-5.2openai/gpt-5.2 | 89.3 | 90% | $1.75 / $14.00 | $4.81 | 2.7 |
Budget champions : 80+ score, cheapest first
| # | Model | Score | % of top score | In / Out per 1M | Blended $/1M | Value |
|---|---|---|---|---|---|---|
| 1 | DeepSeek: DeepSeek V4 Flashdeepseek/deepseek-v4-flash | 80.4 | 81% | $0.10 / $0.20 | $0.12 | 94.1 |
| 2 | DeepSeek: DeepSeek V3deepseek/deepseek-chat | 84.1 | 84% | $0.20 / $0.80 | $0.35 | 34.4 |
| 3 | DeepSeek: DeepSeek V4 Prodeepseek/deepseek-v4-pro | 88.1 | 88% | $0.43 / $0.87 | $0.54 | 23.2 |
| 4 | MoonshotAI: Kimi K2.5moonshotai/kimi-k2.5 | 80.2 | 80% | $0.35 / $1.89 | $0.73 | 15.6 |
| 5 | Z.ai: GLM 5z-ai/glm-5 | 85.0 | 85% | $0.60 / $1.92 | $0.93 | 13.1 |
| 6 | MoonshotAI: Kimi K2.6moonshotai/kimi-k2.6 | 88.9 | 89% | $0.67 / $3.39 | $1.35 | 9.4 |
| 7 | Z.ai: GLM 5.1z-ai/glm-5.1 | 85.7 | 86% | $0.98 / $3.08 | $1.50 | 8.2 |
| 8 | OpenAI: GPT-5openai/gpt-5 | 82.0 | 82% | $1.25 / $10.00 | $3.44 | 3.4 |
| 9 | Google: Gemini 3.1 Pro Previewgoogle/gemini-3.1-pro-preview | 92.4 | 93% | $2.00 / $12.00 | $4.50 | 2.9 |
| 10 | OpenAI: GPT-5.2openai/gpt-5.2 | 89.3 | 90% | $1.75 / $14.00 | $4.81 | 2.7 |
Assumptions
Value = blended benchmark score divided by blended price per million tokens, indexed to the leader. A 3:1 input:output ratio fits most chat and RAG workloads; estimate your exact mix with the cost calculator. Scoring details in the methodology. Models with fewer than 4 independent benchmarks are excluded rather than guessed.