someaistuff.

Benchmark explorer.

LiveBench Overall2.0 · 2026-10-14 (dataset 2026-09-02) · Higher is better
Source ↗
Method

Equal mean of available suites; seven suites in this release. Each suite is a single-run mean.

Score × cost

Where performance meets price

LiveBench · 1M input + 250K output tokens · uncached USD

22 priced resultsPareto frontier
Anthropic: Claude Fable 5.1 ↗87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter

Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.

LiveBench Overall score versus workload cost20%40%60%80%100%$0.01$0.10$1$10$100Higher score is betterWorkload cost · USD · log scaleNVIDIA: Nemotron 3.5 Lightning · 28.69% LiveBench · $0.055 workload · Darkbloom · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.025 input / $0.12 output per 1M tokens · darkbloom/int4 · quantization=int4.DeepSeek: DeepSeek V4.1 Flash · 78.04% LiveBench · $0.0684 workload · Baidu · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0342 input / $0.1368 output per 1M tokens · baidu/fp8 · quantization=fp8.GLM-5.3-Flash · 70.61% LiveBench · $0.0817 workload · OpenInference · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0517 input / $0.12 output per 1M tokens · open-inference/fp4 · quantization=fp4.OpenAI: GPT-6 Luna · 69.4% LiveBench · $0.1125 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.05 input / $0.25 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.DeepSeek-V4.1-Flash-1017 · 79.2% LiveBench · $0.135 workload · Decart · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.09 input / $0.18 output per 1M tokens · decart/fp4 · quantization=fp4.Google: Gemma 4 31B · 41.53% LiveBench · $0.185 workload · CoreWeave · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.10 input / $0.34 output per 1M tokens · coreweave/fp4 · quantization=fp4.Qwen: Qwen3.8 Flash · 68.79% LiveBench · $0.2675 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.15 input / $0.47 output per 1M tokens · alibaba/us-east-1.Qwen3.8-27B · 62.66% LiveBench · $0.3515 workload · Chutes · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.039 input / $1.25 output per 1M tokens · chutes/fp8 · quantization=fp8.DeepSeek-V4-Flash-Vision-Exp · 63.63% LiveBench · $0.3773 workload · DeepInfra · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.2156 input / $0.6468 output per 1M tokens · deepinfra/fp8 · quantization=fp8.GLM-5.3 · 60.74% LiveBench · $0.429 workload · Wafer · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.079 input / $1.40 output per 1M tokens · wafer/us · quantization=fp4.Reflection: Beam 1.0 501B A23B · 55.5% LiveBench · $0.50 workload · Nebius · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.25 input / $1.00 output per 1M tokens · nebius/fp8 · quantization=fp8.Xiaomi: MiMo-V2.6-Pro · 74.87% LiveBench · $0.5828 workload · DeepInfra · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $0.387 input / $0.783 output per 1M tokens · deepinfra/fp8 · quantization=fp8.Step-5-Preview · 62.01% LiveBench · $1.675 workload · DeepInfra · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $2.70 output per 1M tokens.OpenAI: GPT-6.1 Sol · 85.41% LiveBench · $2.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.Google: Gemini 4 Argon · 84.1% LiveBench · $2.25 workload · Google AI Studio · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · google-ai-studio/flex · tier=flex.Meta: Muse Spark Open · 57.57% LiveBench · $2.3125 workload · baseten · via Hugging Face · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.25 input / $4.25 output per 1M tokens.Kimi-K3 · 63.39% LiveBench · $2.575 workload · Sail Research · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.20 input / $5.50 output per 1M tokens · sail-research/fp4/us · quantization=fp4.Qwen: Qwen3.8 2.4T A95B · 64.24% LiveBench · $3.50 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $6.00 output per 1M tokens · alibaba.Anthropic: Claude Sonnet 5.5 · 82.46% LiveBench · $4.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $10.00 output per 1M tokens · amazon-bedrock.Anthropic: Claude Opus 5.5 · 86.53% LiveBench · $9.00 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $4.00 input / $20.00 output per 1M tokens · amazon-bedrock.OpenAI: GPT-6 Astra · 86.14% LiveBench · $11.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $5.00 input / $25.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.Anthropic: Claude Fable 5.1 · 87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.NVIDIA: Nemotron 3.5 Lightning · 28.69% LiveBench · $0.055 workload · Darkbloom · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.025 input / $0.12 output per 1M tokens · darkbloom/int4 · quantization=int4.1DeepSeek: DeepSeek V4.1 Flash · 78.04% LiveBench · $0.0684 workload · Baidu · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0342 input / $0.1368 output per 1M tokens · baidu/fp8 · quantization=fp8.2GLM-5.3-Flash · 70.61% LiveBench · $0.0817 workload · OpenInference · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0517 input / $0.12 output per 1M tokens · open-inference/fp4 · quantization=fp4.3OpenAI: GPT-6 Luna · 69.4% LiveBench · $0.1125 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.05 input / $0.25 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.4DeepSeek-V4.1-Flash-1017 · 79.2% LiveBench · $0.135 workload · Decart · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.09 input / $0.18 output per 1M tokens · decart/fp4 · quantization=fp4.5Google: Gemma 4 31B · 41.53% LiveBench · $0.185 workload · CoreWeave · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.10 input / $0.34 output per 1M tokens · coreweave/fp4 · quantization=fp4.6Qwen: Qwen3.8 Flash · 68.79% LiveBench · $0.2675 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.15 input / $0.47 output per 1M tokens · alibaba/us-east-1.7Qwen3.8-27B · 62.66% LiveBench · $0.3515 workload · Chutes · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.039 input / $1.25 output per 1M tokens · chutes/fp8 · quantization=fp8.8DeepSeek-V4-Flash-Vision-Exp · 63.63% LiveBench · $0.3773 workload · DeepInfra · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.2156 input / $0.6468 output per 1M tokens · deepinfra/fp8 · quantization=fp8.9GLM-5.3 · 60.74% LiveBench · $0.429 workload · Wafer · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.079 input / $1.40 output per 1M tokens · wafer/us · quantization=fp4.10Reflection: Beam 1.0 501B A23B · 55.5% LiveBench · $0.50 workload · Nebius · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.25 input / $1.00 output per 1M tokens · nebius/fp8 · quantization=fp8.11Xiaomi: MiMo-V2.6-Pro · 74.87% LiveBench · $0.5828 workload · DeepInfra · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $0.387 input / $0.783 output per 1M tokens · deepinfra/fp8 · quantization=fp8.12Step-5-Preview · 62.01% LiveBench · $1.675 workload · DeepInfra · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $2.70 output per 1M tokens.13OpenAI: GPT-6.1 Sol · 85.41% LiveBench · $2.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.14Google: Gemini 4 Argon · 84.1% LiveBench · $2.25 workload · Google AI Studio · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · google-ai-studio/flex · tier=flex.15Meta: Muse Spark Open · 57.57% LiveBench · $2.3125 workload · baseten · via Hugging Face · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.25 input / $4.25 output per 1M tokens.16Kimi-K3 · 63.39% LiveBench · $2.575 workload · Sail Research · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.20 input / $5.50 output per 1M tokens · sail-research/fp4/us · quantization=fp4.17Qwen: Qwen3.8 2.4T A95B · 64.24% LiveBench · $3.50 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $6.00 output per 1M tokens · alibaba.18Anthropic: Claude Sonnet 5.5 · 82.46% LiveBench · $4.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $10.00 output per 1M tokens · amazon-bedrock.19Anthropic: Claude Opus 5.5 · 86.53% LiveBench · $9.00 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $4.00 input / $20.00 output per 1M tokens · amazon-bedrock.20OpenAI: GPT-6 Astra · 86.14% LiveBench · $11.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $5.00 input / $25.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.21Anthropic: Claude Fable 5.1 · 87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.22LiveBench Overall score versus workload cost20%40%60%80%100%$0.01$0.10$1$10$100Higher score is betterWorkload cost · USD · log scaleNVIDIA: Nemotron 3.5 Lightning · 28.69% LiveBench · $0.055 workload · Darkbloom · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.025 input / $0.12 output per 1M tokens · darkbloom/int4 · quantization=int4.DeepSeek: DeepSeek V4.1 Flash · 78.04% LiveBench · $0.0684 workload · Baidu · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0342 input / $0.1368 output per 1M tokens · baidu/fp8 · quantization=fp8.GLM-5.3-Flash · 70.61% LiveBench · $0.0817 workload · OpenInference · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0517 input / $0.12 output per 1M tokens · open-inference/fp4 · quantization=fp4.OpenAI: GPT-6 Luna · 69.4% LiveBench · $0.1125 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.05 input / $0.25 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.DeepSeek-V4.1-Flash-1017 · 79.2% LiveBench · $0.135 workload · Decart · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.09 input / $0.18 output per 1M tokens · decart/fp4 · quantization=fp4.Google: Gemma 4 31B · 41.53% LiveBench · $0.185 workload · CoreWeave · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.10 input / $0.34 output per 1M tokens · coreweave/fp4 · quantization=fp4.Qwen: Qwen3.8 Flash · 68.79% LiveBench · $0.2675 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.15 input / $0.47 output per 1M tokens · alibaba/us-east-1.Qwen3.8-27B · 62.66% LiveBench · $0.3515 workload · Chutes · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.039 input / $1.25 output per 1M tokens · chutes/fp8 · quantization=fp8.DeepSeek-V4-Flash-Vision-Exp · 63.63% LiveBench · $0.3773 workload · DeepInfra · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.2156 input / $0.6468 output per 1M tokens · deepinfra/fp8 · quantization=fp8.GLM-5.3 · 60.74% LiveBench · $0.429 workload · Wafer · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.079 input / $1.40 output per 1M tokens · wafer/us · quantization=fp4.Reflection: Beam 1.0 501B A23B · 55.5% LiveBench · $0.50 workload · Nebius · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.25 input / $1.00 output per 1M tokens · nebius/fp8 · quantization=fp8.Xiaomi: MiMo-V2.6-Pro · 74.87% LiveBench · $0.5828 workload · DeepInfra · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $0.387 input / $0.783 output per 1M tokens · deepinfra/fp8 · quantization=fp8.Step-5-Preview · 62.01% LiveBench · $1.675 workload · DeepInfra · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $2.70 output per 1M tokens.OpenAI: GPT-6.1 Sol · 85.41% LiveBench · $2.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.Google: Gemini 4 Argon · 84.1% LiveBench · $2.25 workload · Google AI Studio · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · google-ai-studio/flex · tier=flex.Meta: Muse Spark Open · 57.57% LiveBench · $2.3125 workload · baseten · via Hugging Face · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.25 input / $4.25 output per 1M tokens.Kimi-K3 · 63.39% LiveBench · $2.575 workload · Sail Research · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.20 input / $5.50 output per 1M tokens · sail-research/fp4/us · quantization=fp4.Qwen: Qwen3.8 2.4T A95B · 64.24% LiveBench · $3.50 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $6.00 output per 1M tokens · alibaba.Anthropic: Claude Sonnet 5.5 · 82.46% LiveBench · $4.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $10.00 output per 1M tokens · amazon-bedrock.Anthropic: Claude Opus 5.5 · 86.53% LiveBench · $9.00 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $4.00 input / $20.00 output per 1M tokens · amazon-bedrock.OpenAI: GPT-6 Astra · 86.14% LiveBench · $11.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $5.00 input / $25.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.Anthropic: Claude Fable 5.1 · 87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.NVIDIA: Nemotron 3.5 Lightning · 28.69% LiveBench · $0.055 workload · Darkbloom · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.025 input / $0.12 output per 1M tokens · darkbloom/int4 · quantization=int4.1DeepSeek: DeepSeek V4.1 Flash · 78.04% LiveBench · $0.0684 workload · Baidu · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0342 input / $0.1368 output per 1M tokens · baidu/fp8 · quantization=fp8.2GLM-5.3-Flash · 70.61% LiveBench · $0.0817 workload · OpenInference · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0517 input / $0.12 output per 1M tokens · open-inference/fp4 · quantization=fp4.3OpenAI: GPT-6 Luna · 69.4% LiveBench · $0.1125 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.05 input / $0.25 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.4DeepSeek-V4.1-Flash-1017 · 79.2% LiveBench · $0.135 workload · Decart · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.09 input / $0.18 output per 1M tokens · decart/fp4 · quantization=fp4.5Google: Gemma 4 31B · 41.53% LiveBench · $0.185 workload · CoreWeave · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.10 input / $0.34 output per 1M tokens · coreweave/fp4 · quantization=fp4.6Qwen: Qwen3.8 Flash · 68.79% LiveBench · $0.2675 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.15 input / $0.47 output per 1M tokens · alibaba/us-east-1.7Qwen3.8-27B · 62.66% LiveBench · $0.3515 workload · Chutes · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.039 input / $1.25 output per 1M tokens · chutes/fp8 · quantization=fp8.8DeepSeek-V4-Flash-Vision-Exp · 63.63% LiveBench · $0.3773 workload · DeepInfra · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.2156 input / $0.6468 output per 1M tokens · deepinfra/fp8 · quantization=fp8.9GLM-5.3 · 60.74% LiveBench · $0.429 workload · Wafer · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.079 input / $1.40 output per 1M tokens · wafer/us · quantization=fp4.10Reflection: Beam 1.0 501B A23B · 55.5% LiveBench · $0.50 workload · Nebius · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.25 input / $1.00 output per 1M tokens · nebius/fp8 · quantization=fp8.11Xiaomi: MiMo-V2.6-Pro · 74.87% LiveBench · $0.5828 workload · DeepInfra · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $0.387 input / $0.783 output per 1M tokens · deepinfra/fp8 · quantization=fp8.12Step-5-Preview · 62.01% LiveBench · $1.675 workload · DeepInfra · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $2.70 output per 1M tokens.13OpenAI: GPT-6.1 Sol · 85.41% LiveBench · $2.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.14Google: Gemini 4 Argon · 84.1% LiveBench · $2.25 workload · Google AI Studio · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · google-ai-studio/flex · tier=flex.15Meta: Muse Spark Open · 57.57% LiveBench · $2.3125 workload · baseten · via Hugging Face · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.25 input / $4.25 output per 1M tokens.16Kimi-K3 · 63.39% LiveBench · $2.575 workload · Sail Research · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.20 input / $5.50 output per 1M tokens · sail-research/fp4/us · quantization=fp4.17Qwen: Qwen3.8 2.4T A95B · 64.24% LiveBench · $3.50 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $6.00 output per 1M tokens · alibaba.18Anthropic: Claude Sonnet 5.5 · 82.46% LiveBench · $4.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $10.00 output per 1M tokens · amazon-bedrock.19Anthropic: Claude Opus 5.5 · 86.53% LiveBench · $9.00 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $4.00 input / $20.00 output per 1M tokens · amazon-bedrock.20OpenAI: GPT-6 Astra · 86.14% LiveBench · $11.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $5.00 input / $25.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.21Anthropic: Claude Fable 5.1 · 87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.22LiveBench Overall score versus workload cost20%40%60%80%100%$0.01$0.10$1$10$100Higher score is betterWorkload cost · USD · log scaleNVIDIA: Nemotron 3.5 Lightning · 28.69% LiveBench · $0.055 workload · Darkbloom · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.025 input / $0.12 output per 1M tokens · darkbloom/int4 · quantization=int4.DeepSeek: DeepSeek V4.1 Flash · 78.04% LiveBench · $0.0684 workload · Baidu · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0342 input / $0.1368 output per 1M tokens · baidu/fp8 · quantization=fp8.GLM-5.3-Flash · 70.61% LiveBench · $0.0817 workload · OpenInference · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0517 input / $0.12 output per 1M tokens · open-inference/fp4 · quantization=fp4.OpenAI: GPT-6 Luna · 69.4% LiveBench · $0.1125 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.05 input / $0.25 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.DeepSeek-V4.1-Flash-1017 · 79.2% LiveBench · $0.135 workload · Decart · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.09 input / $0.18 output per 1M tokens · decart/fp4 · quantization=fp4.Google: Gemma 4 31B · 41.53% LiveBench · $0.185 workload · CoreWeave · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.10 input / $0.34 output per 1M tokens · coreweave/fp4 · quantization=fp4.Qwen: Qwen3.8 Flash · 68.79% LiveBench · $0.2675 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.15 input / $0.47 output per 1M tokens · alibaba/us-east-1.Qwen3.8-27B · 62.66% LiveBench · $0.3515 workload · Chutes · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.039 input / $1.25 output per 1M tokens · chutes/fp8 · quantization=fp8.DeepSeek-V4-Flash-Vision-Exp · 63.63% LiveBench · $0.3773 workload · DeepInfra · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.2156 input / $0.6468 output per 1M tokens · deepinfra/fp8 · quantization=fp8.GLM-5.3 · 60.74% LiveBench · $0.429 workload · Wafer · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.079 input / $1.40 output per 1M tokens · wafer/us · quantization=fp4.Reflection: Beam 1.0 501B A23B · 55.5% LiveBench · $0.50 workload · Nebius · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.25 input / $1.00 output per 1M tokens · nebius/fp8 · quantization=fp8.Xiaomi: MiMo-V2.6-Pro · 74.87% LiveBench · $0.5828 workload · DeepInfra · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $0.387 input / $0.783 output per 1M tokens · deepinfra/fp8 · quantization=fp8.Step-5-Preview · 62.01% LiveBench · $1.675 workload · DeepInfra · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $2.70 output per 1M tokens.OpenAI: GPT-6.1 Sol · 85.41% LiveBench · $2.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.Google: Gemini 4 Argon · 84.1% LiveBench · $2.25 workload · Google AI Studio · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · google-ai-studio/flex · tier=flex.Meta: Muse Spark Open · 57.57% LiveBench · $2.3125 workload · baseten · via Hugging Face · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.25 input / $4.25 output per 1M tokens.Kimi-K3 · 63.39% LiveBench · $2.575 workload · Sail Research · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.20 input / $5.50 output per 1M tokens · sail-research/fp4/us · quantization=fp4.Qwen: Qwen3.8 2.4T A95B · 64.24% LiveBench · $3.50 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $6.00 output per 1M tokens · alibaba.Anthropic: Claude Sonnet 5.5 · 82.46% LiveBench · $4.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $10.00 output per 1M tokens · amazon-bedrock.Anthropic: Claude Opus 5.5 · 86.53% LiveBench · $9.00 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $4.00 input / $20.00 output per 1M tokens · amazon-bedrock.OpenAI: GPT-6 Astra · 86.14% LiveBench · $11.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $5.00 input / $25.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.Anthropic: Claude Fable 5.1 · 87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.NVIDIA: Nemotron 3.5 Lightning · 28.69% LiveBench · $0.055 workload · Darkbloom · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.025 input / $0.12 output per 1M tokens · darkbloom/int4 · quantization=int4.1DeepSeek: DeepSeek V4.1 Flash · 78.04% LiveBench · $0.0684 workload · Baidu · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0342 input / $0.1368 output per 1M tokens · baidu/fp8 · quantization=fp8.2GLM-5.3-Flash · 70.61% LiveBench · $0.0817 workload · OpenInference · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.0517 input / $0.12 output per 1M tokens · open-inference/fp4 · quantization=fp4.3OpenAI: GPT-6 Luna · 69.4% LiveBench · $0.1125 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.05 input / $0.25 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.4DeepSeek-V4.1-Flash-1017 · 79.2% LiveBench · $0.135 workload · Decart · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.09 input / $0.18 output per 1M tokens · decart/fp4 · quantization=fp4.5Google: Gemma 4 31B · 41.53% LiveBench · $0.185 workload · CoreWeave · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.10 input / $0.34 output per 1M tokens · coreweave/fp4 · quantization=fp4.6Qwen: Qwen3.8 Flash · 68.79% LiveBench · $0.2675 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.15 input / $0.47 output per 1M tokens · alibaba/us-east-1.7Qwen3.8-27B · 62.66% LiveBench · $0.3515 workload · Chutes · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.039 input / $1.25 output per 1M tokens · chutes/fp8 · quantization=fp8.8DeepSeek-V4-Flash-Vision-Exp · 63.63% LiveBench · $0.3773 workload · DeepInfra · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.2156 input / $0.6468 output per 1M tokens · deepinfra/fp8 · quantization=fp8.9GLM-5.3 · 60.74% LiveBench · $0.429 workload · Wafer · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.079 input / $1.40 output per 1M tokens · wafer/us · quantization=fp4.10Reflection: Beam 1.0 501B A23B · 55.5% LiveBench · $0.50 workload · Nebius · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $0.25 input / $1.00 output per 1M tokens · nebius/fp8 · quantization=fp8.11Xiaomi: MiMo-V2.6-Pro · 74.87% LiveBench · $0.5828 workload · DeepInfra · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $0.387 input / $0.783 output per 1M tokens · deepinfra/fp8 · quantization=fp8.12Step-5-Preview · 62.01% LiveBench · $1.675 workload · DeepInfra · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $2.70 output per 1M tokens.13OpenAI: GPT-6.1 Sol · 85.41% LiveBench · $2.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.14Google: Gemini 4 Argon · 84.1% LiveBench · $2.25 workload · Google AI Studio · via OpenRouter · Evaluation: Effort: high · single run · equal mean of 7 suites. Price: $1.00 input / $5.00 output per 1M tokens · google-ai-studio/flex · tier=flex.15Meta: Muse Spark Open · 57.57% LiveBench · $2.3125 workload · baseten · via Hugging Face · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.25 input / $4.25 output per 1M tokens.16Kimi-K3 · 63.39% LiveBench · $2.575 workload · Sail Research · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $1.20 input / $5.50 output per 1M tokens · sail-research/fp4/us · quantization=fp4.17Qwen: Qwen3.8 2.4T A95B · 64.24% LiveBench · $3.50 workload · Alibaba · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $6.00 output per 1M tokens · alibaba.18Anthropic: Claude Sonnet 5.5 · 82.46% LiveBench · $4.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $2.00 input / $10.00 output per 1M tokens · amazon-bedrock.19Anthropic: Claude Opus 5.5 · 86.53% LiveBench · $9.00 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $4.00 input / $20.00 output per 1M tokens · amazon-bedrock.20OpenAI: GPT-6 Astra · 86.14% LiveBench · $11.25 workload · OpenAI · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $5.00 input / $25.00 output per 1M tokens · openai/flex · tier=flex · prompt ≤ 272000 tokens.21Anthropic: Claude Fable 5.1 · 87.99% LiveBench · $22.50 workload · Amazon Bedrock · via OpenRouter · Evaluation: Effort: max · single run · equal mean of 7 suites. Price: $10.00 input / $50.00 output per 1M tokens · amazon-bedrock.22

Numbers match the model key. Shared positions show an index range.

Model key Score / workload

  1. NVIDIA: Nemotron 3.5 Lightning28.69% / $0.055
  2. DeepSeek: DeepSeek V4.1 Flash78.04% / $0.0684
  3. GLM-5.3-Flash70.61% / $0.0817
  4. OpenAI: GPT-6 Luna69.4% / $0.1125
  5. DeepSeek-V4.1-Flash-101779.2% / $0.135
  6. Google: Gemma 4 31B41.53% / $0.185
  7. Qwen: Qwen3.8 Flash68.79% / $0.2675
  8. Qwen3.8-27B62.66% / $0.3515
  9. DeepSeek-V4-Flash-Vision-Exp63.63% / $0.3773
  10. GLM-5.360.74% / $0.429
  11. Reflection: Beam 1.0 501B A23B55.5% / $0.50
  12. Xiaomi: MiMo-V2.6-Pro74.87% / $0.5828
  13. Step-5-Preview62.01% / $1.675
  14. OpenAI: GPT-6.1 Sol85.41% / $2.25
  15. Google: Gemini 4 Argon84.1% / $2.25
  16. Meta: Muse Spark Open57.57% / $2.3125
  17. Kimi-K363.39% / $2.575
  18. Qwen: Qwen3.8 2.4T A95B64.24% / $3.50
  19. Anthropic: Claude Sonnet 5.582.46% / $4.50
  20. Anthropic: Claude Opus 5.586.53% / $9.00
  21. OpenAI: GPT-6 Astra86.14% / $11.25
  22. Anthropic: Claude Fable 5.187.99% / $22.50

Pareto frontier

Frontier: no result is cheaper with an equal or better score, or equally priced with a better score. Each point uses its lowest generally available workload price from one endpoint; evaluation and provider configurations remain visible.

49 results · percent

Page 1 / 1
LiveBench Overall results, configurations and model prices
#ModelLiveBenchConfigurationAPI rate / unitSource
1Anthropic: Claude Fable 5.1anthropic87.99%Effort: max · single run · equal mean of 7 suitesIndependent result$10.00 in / $50.00 out/ 1M tokens · Amazon Bedrock · via OpenRouteramazon-bedrockResult ↗Oct 18, 2026
2Anthropic: Claude Opus 5.5anthropic86.53%Effort: max · single run · equal mean of 7 suitesIndependent result$4.00 in / $20.00 out/ 1M tokens · Amazon Bedrock · via OpenRouteramazon-bedrockResult ↗Oct 18, 2026
3OpenAI: GPT-6 Astraopenai86.14%Effort: max · single run · equal mean of 7 suitesIndependent result$5.00 in / $25.00 out/ 1M tokens · OpenAI · via OpenRouteropenai/flex · tier=flex · prompt ≤ 272000 tokensResult ↗Oct 18, 2026
4OpenAI: GPT-6.1 Solopenai85.41%Effort: max · single run · equal mean of 7 suitesIndependent result$1.00 in / $5.00 out/ 1M tokens · OpenAI · via OpenRouteropenai/flex · tier=flex · prompt ≤ 272000 tokensResult ↗Oct 18, 2026
5Google: Gemini 4 Argongoogle84.1%Effort: high · single run · equal mean of 7 suitesIndependent result$1.00 in / $5.00 out/ 1M tokens · Google AI Studio · via OpenRoutergoogle-ai-studio/flex · tier=flexResult ↗Oct 18, 2026
6Claude Fable 5Anthropic83.61%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
7Anthropic: Claude Sonnet 5.5anthropic82.46%Effort: max · single run · equal mean of 7 suitesIndependent result$2.00 in / $10.00 out/ 1M tokens · Amazon Bedrock · via OpenRouteramazon-bedrockResult ↗Oct 18, 2026
8Claude Opus 5Anthropic82.09%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
9GPT-6 SolOpenAI81.91%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
10DeepSeek-V4.1-Flash-1017deepseek-ai79.2%Effort: max · single run · equal mean of 7 suitesIndependent result$0.07 in / $0.50 out/ 1M tokens · Relace · via OpenRouterrelaceResult ↗Oct 18, 2026
11Muse Wonder 1.0Meta78.1%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
12DeepSeek: DeepSeek V4.1 Flashdeepseek78.04%Effort: max · single run · equal mean of 7 suitesIndependent result$0.0239 in / $0.375 out/ 1M tokens · OpenInference · via OpenRouteropen-inference/fp4 · quantization=fp4Result ↗Oct 18, 2026
13Claude Opus 4.8Anthropic76.77%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
14GPT-5.6 SolOpenAI76.26%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
15GPT-5.6 TerraOpenAI74.99%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
16Xiaomi: MiMo-V2.6-Proxiaomi74.87%Effort: high · single run · equal mean of 7 suitesIndependent result$0.387 in / $0.783 out/ 1M tokens · DeepInfra · via OpenRouterdeepinfra/fp8 · quantization=fp8Result ↗Oct 18, 2026
17Muse Spark 1.3Meta74.76%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
18Grok 4.6xAI74.64%Effort: xhigh · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
19Claude Sonnet 5Anthropic74.63%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
20Claude Haiku 5.5Anthropic74.49%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
21Grok 4.7xAI74.09%Effort: xhigh · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
22GLM-5.3-Flashzai-org70.61%Effort: max · single run · equal mean of 7 suitesIndependent result$0.0517 in / $0.12 out/ 1M tokens · OpenInference · via OpenRouteropen-inference/fp4 · quantization=fp4Result ↗Oct 18, 2026
23Grok 4.5xAI70.06%Effort: xhigh · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
24OpenAI: GPT-6 Lunaopenai69.4%Effort: max · single run · equal mean of 7 suitesIndependent result$0.05 in / $0.25 out/ 1M tokens · OpenAI · via OpenRouteropenai/flex · tier=flex · prompt ≤ 272000 tokensResult ↗Oct 18, 2026
25Qwen: Qwen3.8 Flashqwen68.79%Effort: max · single run · equal mean of 7 suitesIndependent result$0.15 in / $0.47 out/ 1M tokens · Alibaba · via OpenRouteralibaba/us-east-1Result ↗Oct 18, 2026
26GPT-5.5OpenAI68.7%Effort: xhigh · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
27Gemini 3.8 FlashGoogle67.34%Effort: high · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
28Gemini 3.7 FlashGoogle67.19%Effort: high · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
29GPT-5.6 LunaOpenAI66.09%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
30Claude Sonnet 4.6Anthropic65.09%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
31Qwen: Qwen3.8 2.4T A95Bqwen64.24%Effort: max · single run · equal mean of 7 suitesIndependent result$2.00 in / $6.00 out/ 1M tokens · Alibaba · via OpenRouteralibabaResult ↗Oct 18, 2026
32DeepSeek-V4-Flash-Vision-Expdeepseek-ai63.63%Effort: max · single run · equal mean of 7 suitesIndependent result$0.2156 in / $0.6468 out/ 1M tokens · DeepInfra · via OpenRouterdeepinfra/fp8 · quantization=fp8Result ↗Oct 18, 2026
33Kimi-K3moonshotai63.39%Effort: max · single run · equal mean of 7 suitesIndependent result$0.238 in / $16.00 out/ 1M tokens · Wafer · via OpenRouterwafer · quantization=mxfp4Result ↗Oct 18, 2026
34Qwen3.8-27BQwen62.66%Effort: max · single run · equal mean of 7 suitesIndependent result$0.039 in / $1.25 out/ 1M tokens · Chutes · via OpenRouterchutes/fp8 · quantization=fp8Result ↗Oct 18, 2026
35Muse Spark 1.2Meta62.4%Effort: xhigh · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
36Step-5-Previewstepfun-ai62.01%Effort: max · single run · equal mean of 7 suitesIndependent result$1.00 in / $2.70 out/ 1M tokens · DeepInfraResult ↗Oct 18, 2026
37GLM-5.3zai-org60.74%Effort: max · single run · equal mean of 7 suitesIndependent result$0.079 in / $1.40 out/ 1M tokens · Wafer · via OpenRouterwafer/us · quantization=fp4Result ↗Oct 18, 2026
38Meta: Muse Spark Openmeta57.57%Effort: max · single run · equal mean of 7 suitesIndependent result$1.25 in / $4.25 out/ 1M tokens · baseten · via Hugging FaceResult ↗Oct 18, 2026
39smaug-flashUnknown57.54%Effort not specified · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
40Reflection: Beam 1.0 501B A23Breflection55.5%Effort: max · single run · equal mean of 7 suitesIndependent result$0.25 in / $1.00 out/ 1M tokens · Nebius · via OpenRouternebius/fp8 · quantization=fp8Result ↗Oct 18, 2026
41Gemini 3.1 ProGoogle54.91%Effort: high · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
42MiniMax M3MiniMax53.74%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
43InklingThinking Machines51.91%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
44Nemotron 3.5 Super VLNVIDIA51.47%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
45Mistral Large 4Mistral AI47.7%Effort: high · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
46Claude Haiku 4.5Anthropic46.04%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
47Google: Gemma 4 31Bgoogle41.53%Effort: max · single run · equal mean of 7 suitesIndependent result$0.10 in / $0.34 out/ 1M tokens · CoreWeave · via OpenRoutercoreweave/fp4 · quantization=fp4Result ↗Oct 18, 2026
48Nemotron 3 UltraNVIDIA40.83%Effort: max · single run · equal mean of 7 suitesIndependent resultSee provider quotesResult ↗Oct 18, 2026
49NVIDIA: Nemotron 3.5 Lightningnvidia28.69%Effort: max · single run · equal mean of 7 suitesIndependent result$0.025 in / $0.12 out/ 1M tokens · Darkbloom · via OpenRouterdarkbloom/int4 · quantization=int4Result ↗Oct 18, 2026