Category intelligence / models
Model efficiency
Capability relative to token price, context, latency, and throughput.
Evidence catalog
Open category terminal →Live Efficiency tables
| Table | Coverage | Confidence | Freshness | Cadence | Updated | Open |
|---|---|---|---|---|---|---|
| Quality per dollarIntelligence points per weighted inference dollar at public API prices. | 100 | Source | Current | 6-hour refresh | Jul 28, 2026 | View table → |
| Context per dollarUsable context capacity relative to published input-token pricing. | 100 | Source | Current | 6-hour refresh | Jul 28, 2026 | View table → |
| LiveBench cost efficiencyOfficial cost per successful LiveBench task beside capability, token use, and the exact prices frozen into the release. | 45 | 100% | current | Daily verification · semiannual release | Jun 25, 2026 | View table → |
MODEL ECONOMICS
Quality per dollar
| # | Model | Points / $ | Input / 1M | Output / 1M | Context |
|---|---|---|---|---|---|
| 01 | InclusionaiLing-2.6-flash | 564.00 | $0.01 | $0.03 | 262K |
| 02 | TencentHy3 preview | 237.81 | $0.06 | $0.21 | 262K |
| 03 | OpenAIgpt-oss-120b | 174.04 | $0.04 | $0.17 | 131K |
| 04 | DeepseekDeepSeek V4 Flash | 164.49 | $0.14 | $0.28 | 1.0M |
| 05 | QwenQwen3.5-9B | 155.64 | $0.10 | $0.15 | 262K |
| 06 | XiaomiMiMo-V2.5 | 151.84 | $0.14 | $0.28 | 1.1M |
| 07 | OpenAIgpt-oss-20b | 132.44 | $0.03 | $0.14 | 131K |
| 08 | MetaLlama 3.1 8B Instruct | 104.83 | $0.05 | $0.08 | 131K |
| 09 | GoogleGemma 4 31B | 87.76 | $0.14 | $0.40 | 262K |
| 10 | NvidiaNemotron 3 Nano 30B A3B | 87.38 | $0.05 | $0.20 | 262K |
| 11 | DeepseekDeepSeek V3.2 | 87.13 | $0.27 | $0.40 | 164K |
| 12 | NvidiaNemotron 3 Super | 79.07 | $0.08 | $0.40 | 1M |
| 13 | GoogleGemma 4 26B A4B | 73.43 | $0.14 | $0.42 | 262K |
| 14 | MistralMinistral 3 3B 2512 | 65.00 | $0.10 | $0.10 | 131K |
| 15 | InclusionaiRing-2.6-1T | 62.77 | $0.07 | $0.63 | 262K |
| 16 | MistralMinistral 3 8B 2512 | 60.00 | $0.15 | $0.15 | 262K |
| 17 | DeepseekDeepSeek V4 Pro | 58.19 | $0.43 | $0.87 | 1.0M |
| 18 | MistralMinistral 3 14B 2512 | 55.50 | $0.20 | $0.20 | 262K |
| 19 | XiaomiMiMo-V2.5-Pro | 55.44 | $0.43 | $0.87 | 1.1M |
| 20 | Nex AgiNex-N2-Pro | 50.46 | $0.25 | $1.00 | 262K |
| 21 | QwenQwen3 32B | 50.00 | $0.08 | $0.28 | 131K |
| 22 | MinimaxMiniMax M2.7 | 46.89 | $0.25 | $1.00 | 205K |
| 23 | MinimaxMiniMax M3 | 45.54 | $0.30 | $1.20 | 1.0M |
| 24 | GoogleGemma 3 12B | 44.00 | $0.05 | $0.15 | 131K |
| 25 | QwenQwen3.6 35B A3B | 40.25 | $0.14 | $1.00 | 262K |
| 26 | MistralMistral Small 4 | 40.21 | $0.15 | $0.60 | 262K |
| 27 | MetaLlama 4 Scout | 40.00 | $0.10 | $0.30 | 1.3M |
| 28 | OpenAIGPT-5.4 Nano | 38.68 | $0.20 | $1.25 | 400K |
| 29 | QwenQwen3.7 Plus | 37.50 | $0.32 | $1.28 | 1M |
| 30 | DeepseekDeepSeek V3.1 Terminus | 37.19 | $0.27 | $1.00 | 164K |
| 31 | KwaipilotKAT-Coder-Pro V2 | 34.56 | $0.30 | $1.20 | 262K |
| 32 | InceptionMercury 2 | 34.24 | $0.25 | $0.75 | 128K |
| 33 | QwenQwen3 Coder Next | 33.63 | $0.11 | $0.80 | 262K |
| 34 | StepfunStep 3.7 Flash | 33.21 | $0.20 | $1.15 | 262K |
| 35 | QwenQwen3.5-35B-A3B | 30.48 | $0.15 | $1.00 | 262K |
| 36 | OpenAIGPT-4.1 Nano | 29.54 | $0.10 | $0.40 | 1.0M |
| 37 | UpstageSolar Pro 3 | 28.92 | $0.15 | $0.60 | 128K |
| 38 | MetaLlama 3.3 70B Instruct | 28.27 | $0.13 | $0.40 | 131K |
| 39 | QwenQwen3 Next 80B A3B Thinking | 27.41 | $0.10 | $0.78 | 262K |
| 40 | Z AiGLM 5.2 | 26.77 | $0.73 | $2.30 | 1.0M |
| 41 | Arcee AiTrinity Large Thinking | 26.28 | $0.22 | $0.85 | 262K |
| 42 | QwenQwen3.6 Plus | 25.65 | $0.33 | $1.95 | 1M |
| 43 | Z AiGLM 4.7 | 23.86 | $0.40 | $1.75 | 205K |
| 44 | QwenQwen3.6 27B | 23.56 | $0.30 | $2.00 | 262K |
| 45 | xAIGrok Build 0.1 | 22.74 | $1.00 | $2.00 | 256K |
| 46 | QwenQwen3 8B | 22.40 | $0.12 | $0.45 | 131K |
| 47 | MetaLlama 4 Maverick | 22.00 | $0.20 | $0.80 | 1.0M |
| 48 | OpenAIGPT-5.6 Luna | 21.56 | $0.50 | $3.00 | 1.1M |
| 49 | GoogleGemini 3.1 Flash Lite Preview | 21.05 | $0.25 | $1.50 | 1.0M |
| 50 | GoogleGemma 3 27B | 20.70 | $0.08 | $0.45 | 262K |
CONTEXT ECONOMICS
Context per dollar
| # | Model | M ctx / $ | Input / 1M | Output / 1M | Context |
|---|---|---|---|---|---|
| 01 | InclusionaiLing-2.6-flash | 26.21 | $0.01 | $0.03 | 262K |
| 02 | MetaLlama 4 Scout | 13.11 | $0.10 | $0.30 | 1.3M |
| 03 | NvidiaNemotron 3 Super | 11.76 | $0.08 | $0.40 | 1M |
| 04 | OpenAIGPT-4.1 Nano | 10.48 | $0.10 | $0.40 | 1.0M |
| 05 | XiaomiMiMo-V2.5 | 7.50 | $0.14 | $0.28 | 1.1M |
| 06 | DeepseekDeepSeek V4 Flash | 7.49 | $0.14 | $0.28 | 1.0M |
| 07 | MetaLlama 4 Maverick | 5.24 | $0.20 | $0.80 | 1.0M |
| 08 | NvidiaNemotron 3 Nano 30B A3B | 5.24 | $0.05 | $0.20 | 262K |
| 09 | OpenAIgpt-oss-20b | 4.37 | $0.03 | $0.14 | 131K |
| 10 | GoogleGemini 3.1 Flash Lite Preview | 4.19 | $0.25 | $1.50 | 1.0M |
| 11 | TencentHy3 preview | 4.16 | $0.06 | $0.21 | 262K |
| 12 | OpenAIgpt-oss-120b | 3.54 | $0.04 | $0.17 | 131K |
| 13 | MinimaxMiniMax M3 | 3.50 | $0.30 | $1.20 | 1.0M |
| 14 | GoogleGemini 3.5 Flash Lite | 3.50 | $0.30 | $2.50 | 1.0M |
| 15 | InclusionaiRing-2.6-1T | 3.50 | $0.07 | $0.63 | 262K |
| 16 | AmazonNova 2 Lite | 3.33 | $0.30 | $2.50 | 1M |
| 17 | GoogleGemma 3 27B | 3.28 | $0.08 | $0.45 | 262K |
| 18 | QwenQwen3.7 Plus | 3.13 | $0.32 | $1.28 | 1M |
| 19 | QwenQwen3.6 Plus | 3.08 | $0.33 | $1.95 | 1M |
| 20 | QwenQwen3 Next 80B A3B Thinking | 2.69 | $0.10 | $0.78 | 262K |
| 21 | QwenQwen3.5-9B | 2.62 | $0.10 | $0.15 | 262K |
| 22 | MetaLlama 3.1 8B Instruct | 2.62 | $0.05 | $0.08 | 131K |
| 23 | GoogleGemma 3 12B | 2.62 | $0.05 | $0.15 | 131K |
| 24 | OpenAIGPT-4.1 Mini | 2.62 | $0.40 | $1.60 | 1.0M |
| 25 | XiaomiMiMo-V2.5-Pro | 2.41 | $0.43 | $0.87 | 1.1M |
| 26 | DeepseekDeepSeek V4 Pro | 2.41 | $0.43 | $0.87 | 1.0M |
| 27 | QwenQwen3 Coder Next | 2.38 | $0.11 | $0.80 | 262K |
| 28 | OpenAIGPT-5.6 Luna | 2.10 | $0.50 | $3.00 | 1.1M |
| 29 | OpenAIGPT-5.4 Nano | 2.00 | $0.20 | $1.25 | 400K |
| 30 | QwenQwen3.6 35B A3B | 1.87 | $0.14 | $1.00 | 262K |
| 31 | GoogleGemma 4 31B | 1.87 | $0.14 | $0.40 | 262K |
| 32 | GoogleGemma 4 26B A4B | 1.87 | $0.14 | $0.42 | 262K |
| 33 | QwenQwen3.5-35B-A3B | 1.75 | $0.15 | $1.00 | 262K |
| 34 | MistralMistral Small 4 | 1.75 | $0.15 | $0.60 | 262K |
| 35 | MistralMinistral 3 8B 2512 | 1.75 | $0.15 | $0.15 | 262K |
| 36 | QwenQwen3 32B | 1.64 | $0.08 | $0.28 | 131K |
| 37 | OpenAIGPT-5 Mini | 1.60 | $0.25 | $2.00 | 400K |
| 38 | Z AiGLM 5.2 | 1.43 | $0.73 | $2.30 | 1.0M |
| 39 | StepfunStep 3.7 Flash | 1.31 | $0.20 | $1.15 | 262K |
| 40 | MistralMinistral 3 14B 2512 | 1.31 | $0.20 | $0.20 | 262K |
| 41 | MistralMinistral 3 3B 2512 | 1.31 | $0.10 | $0.10 | 131K |
| 42 | Arcee AiTrinity Large Thinking | 1.19 | $0.22 | $0.85 | 262K |
| 43 | QwenQwen3 8B | 1.12 | $0.12 | $0.45 | 131K |
| 44 | Nex AgiNex-N2-Pro | 1.05 | $0.25 | $1.00 | 262K |
| 45 | ThinkingmachinesInkling | 1.05 | $1.00 | $4.05 | 1.0M |
| 46 | QwenQwen3.5-122B-A10B | 1.01 | $0.26 | $2.08 | 262K |
| 47 | MetaLlama 3.3 70B Instruct | 1.01 | $0.13 | $0.40 | 131K |
| 48 | QwenQwen3.6 27B | 0.87 | $0.30 | $2.00 | 262K |
| 49 | KwaipilotKAT-Coder-Pro V2 | 0.87 | $0.30 | $1.20 | 262K |
| 50 | QwenQwen3 235B A22B Thinking 2507 | 0.87 | $0.30 | $3.00 | 262K |
Use the public category as your baseline
Analyze this category →Browse every live table →